1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235
2236
2237
2238
2239
2240
2241
2242
2243
2244
2245
2246
2247
2248
2249
2250
2251
2252
2253
2254
2255
2256
2257
2258
2259
2260
2261
2262
2263
2264
2265
2266
2267
2268
2269
2270
2271
2272
2273
2274
2275
2276
2277
2278
2279
2280
2281
2282
2283
2284
2285
2286
2287
2288
2289
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
2304
2305
2306
2307
2308
2309
2310
2311
2312
2313
2314
2315
2316
2317
2318
2319
2320
2321
2322
2323
2324
2325
2326
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336
2337
2338
2339
2340
2341
2342
2343
2344
2345
2346
2347
2348
2349
2350
2351
2352
2353
2354
2355
2356
2357
2358
2359
2360
2361
2362
2363
2364
2365
2366
2367
2368
2369
2370
2371
2372
2373
2374
2375
2376
2377
2378
2379
2380
2381
2382
2383
2384
2385
2386
2387
2388
2389
2390
2391
2392
2393
2394
2395
2396
2397
2398
2399
2400
2401
2402
2403
2404
2405
2406
2407
2408
2409
2410
2411
2412
2413
2414
2415
2416
2417
2418
2419
2420
2421
2422
2423
2424
2425
2426
2427
2428
2429
2430
2431
2432
2433
2434
2435
2436
2437
2438
2439
2440
2441
2442
2443
2444
2445
2446
2447
2448
2449
2450
2451
2452
2453
2454
2455
2456
2457
2458
2459
2460
2461
2462
2463
2464
2465
2466
2467
2468
2469
2470
2471
2472
2473
2474
2475
2476
2477
2478
2479
2480
2481
2482
2483
2484
2485
2486
2487
2488
2489
2490
2491
2492
2493
2494
2495
2496
2497
2498
2499
2500
2501
2502
2503
2504
2505
2506
2507
2508
2509
2510
2511
2512
2513
2514
2515
2516
2517
2518
2519
2520
2521
2522
2523
2524
2525
2526
2527
2528
2529
2530
2531
2532
2533
2534
2535
2536
2537
2538
2539
2540
2541
2542
2543
2544
2545
2546
2547
2548
2549
2550
2551
2552
2553
2554
2555
2556
2557
2558
2559
2560
2561
2562
2563
2564
2565
2566
2567
2568
2569
2570
2571
2572
2573
2574
2575
2576
2577
2578
2579
2580
2581
2582
2583
2584
2585
2586
2587
2588
2589
2590
2591
2592
2593
2594
2595
2596
2597
2598
2599
2600
2601
2602
2603
2604
2605
2606
2607
2608
2609
2610
2611
2612
2613
2614
2615
2616
2617
2618
2619
2620
2621
2622
2623
2624
2625
2626
2627
2628
2629
2630
2631
2632
2633
2634
2635
2636
2637
2638
2639
2640
2641
2642
2643
2644
2645
2646
2647
2648
2649
2650
2651
2652
2653
2654
2655
2656
2657
2658
2659
2660
2661
2662
2663
2664
2665
2666
2667
2668
2669
2670
2671
2672
2673
2674
2675
2676
2677
2678
2679
2680
2681
2682
2683
2684
2685
2686
2687
2688
2689
2690
2691
2692
2693
2694
2695
2696
2697
2698
2699
2700
2701
2702
2703
2704
2705
2706
2707
2708
2709
2710
2711
2712
2713
2714
2715
2716
2717
2718
2719
2720
2721
2722
2723
2724
2725
2726
2727
2728
2729
2730
2731
2732
2733
2734
2735
2736
2737
2738
2739
2740
2741
2742
2743
2744
2745
2746
2747
2748
2749
2750
2751
2752
2753
2754
2755
2756
2757
2758
2759
2760
2761
2762
2763
2764
2765
2766
2767
2768
2769
2770
2771
2772
2773
2774
2775
2776
2777
2778
2779
2780
2781
2782
2783
2784
2785
2786
2787
2788
2789
2790
2791
2792
2793
2794
2795
2796
2797
2798
2799
2800
2801
2802
2803
2804
2805
2806
2807
2808
2809
2810
2811
2812
2813
2814
2815
2816
2817
2818
2819
2820
2821
2822
2823
2824
2825
2826
2827
2828
2829
2830
2831
2832
2833
2834
2835
2836
2837
2838
2839
2840
2841
2842
2843
2844
2845
2846
2847
2848
2849
2850
2851
2852
2853
2854
2855
2856
2857
2858
2859
2860
2861
2862
2863
2864
2865
2866
2867
2868
2869
2870
2871
2872
2873
2874
2875
2876
2877
2878
2879
2880
2881
2882
2883
2884
2885
2886
2887
2888
2889
2890
2891
2892
2893
2894
2895
2896
2897
2898
2899
2900
2901
2902
2903
2904
2905
2906
2907
2908
2909
2910
2911
2912
2913
2914
2915
2916
2917
2918
2919
2920
2921
2922
2923
2924
2925
2926
2927
2928
2929
2930
2931
2932
2933
2934
2935
2936
2937
2938
2939
2940
2941
2942
2943
2944
2945
2946
2947
2948
2949
2950
2951
2952
2953
2954
2955
2956
2957
2958
2959
2960
2961
2962
2963
2964
2965
2966
2967
2968
2969
2970
2971
2972
2973
2974
2975
2976
2977
2978
2979
2980
2981
2982
2983
2984
2985
2986
2987
2988
2989
2990
2991
2992
2993
2994
2995
2996
2997
2998
2999
3000
3001
3002
3003
3004
3005
3006
3007
3008
3009
3010
3011
3012
3013
3014
3015
3016
3017
3018
3019
3020
3021
3022
3023
3024
3025
3026
3027
3028
3029
3030
3031
3032
3033
3034
3035
3036
3037
3038
3039
3040
3041
3042
3043
3044
3045
3046
3047
3048
3049
3050
3051
3052
3053
3054
3055
3056
3057
3058
3059
3060
3061
3062
3063
3064
3065
3066
3067
3068
3069
3070
3071
3072
3073
3074
3075
3076
3077
3078
3079
3080
3081
3082
3083
3084
3085
3086
3087
3088
3089
3090
3091
3092
3093
3094
3095
3096
3097
3098
3099
3100
3101
3102
3103
3104
3105
3106
3107
3108
3109
3110
3111
3112
3113
3114
3115
3116
3117
3118
3119
3120
3121
3122
3123
3124
3125
3126
3127
3128
3129
3130
3131
3132
3133
3134
3135
3136
3137
3138
3139
3140
3141
3142
3143
3144
3145
3146
3147
3148
3149
3150
3151
3152
3153
3154
3155
3156
3157
3158
3159
3160
3161
3162
3163
3164
3165
3166
3167
3168
3169
3170
3171
3172
3173
3174
3175
3176
3177
3178
3179
3180
3181
3182
3183
3184
3185
3186
3187
3188
3189
3190
3191
3192
3193
3194
3195
3196
3197
3198
3199
3200
3201
3202
3203
3204
3205
3206
3207
3208
3209
3210
3211
3212
3213
3214
3215
3216
3217
3218
3219
3220
3221
3222
3223
3224
3225
3226
3227
3228
3229
3230
3231
3232
3233
3234
3235
3236
3237
3238
3239
3240
3241
3242
3243
3244
3245
3246
3247
3248
3249
3250
3251
3252
3253
3254
3255
3256
3257
3258
3259
3260
3261
3262
3263
3264
3265
3266
3267
3268
3269
3270
3271
3272
3273
3274
3275
3276
3277
3278
3279
3280
3281
3282
3283
3284
3285
3286
3287
3288
3289
3290
3291
3292
3293
3294
3295
3296
3297
3298
3299
3300
3301
3302
3303
3304
3305
3306
3307
3308
3309
3310
3311
3312
3313
3314
3315
3316
3317
3318
3319
3320
3321
3322
3323
3324
3325
3326
3327
3328
3329
3330
3331
3332
3333
3334
3335
3336
3337
3338
3339
3340
3341
3342
3343
3344
3345
3346
3347
3348
3349
3350
3351
3352
3353
3354
3355
3356
3357
3358
3359
3360
3361
3362
3363
3364
3365
3366
3367
3368
3369
3370
3371
3372
3373
3374
3375
3376
3377
3378
3379
3380
3381
3382
3383
3384
3385
3386
3387
3388
3389
3390
3391
3392
3393
3394
3395
3396
3397
3398
3399
3400
3401
3402
3403
3404
3405
3406
3407
3408
3409
3410
3411
3412
3413
3414
3415
3416
3417
3418
3419
3420
3421
3422
3423
3424
3425
3426
3427
3428
3429
3430
3431
3432
3433
3434
3435
3436
3437
3438
3439
3440
3441
3442
3443
3444
3445
3446
3447
3448
3449
3450
3451
3452
3453
3454
3455
3456
3457
3458
3459
3460
3461
3462
3463
3464
3465
3466
3467
3468
3469
3470
3471
3472
3473
3474
3475
3476
3477
3478
3479
3480
3481
3482
3483
3484
3485
3486
3487
3488
3489
3490
3491
3492
3493
3494
3495
3496
3497
3498
3499
3500
3501
3502
3503
3504
3505
3506
3507
3508
3509
3510
3511
3512
3513
3514
3515
3516
3517
3518
3519
3520
3521
3522
3523
3524
3525
3526
3527
3528
3529
3530
3531
3532
3533
3534
3535
3536
3537
3538
3539
3540
3541
3542
3543
3544
3545
3546
3547
3548
3549
3550
3551
3552
3553
3554
3555
3556
3557
3558
3559
3560
3561
3562
3563
3564
3565
3566
3567
3568
3569
3570
3571
3572
3573
3574
3575
3576
3577
3578
3579
3580
3581
3582
3583
3584
3585
3586
3587
3588
3589
3590
3591
3592
3593
3594
3595
3596
3597
3598
3599
3600
3601
3602
3603
3604
3605
3606
3607
3608
3609
3610
3611
3612
3613
3614
3615
3616
3617
3618
3619
3620
3621
3622
3623
3624
3625
3626
3627
3628
3629
3630
3631
3632
3633
3634
3635
3636
3637
3638
3639
3640
3641
3642
3643
3644
3645
3646
3647
3648
3649
3650
3651
3652
3653
3654
3655
3656
3657
3658
3659
3660
3661
3662
3663
3664
3665
3666
3667
3668
3669
3670
3671
3672
3673
3674
3675
3676
3677
3678
3679
3680
3681
3682
3683
3684
3685
3686
3687
3688
3689
3690
3691
3692
3693
3694
3695
3696
3697
3698
3699
3700
3701
3702
3703
3704
3705
3706
3707
3708
3709
3710
3711
3712
3713
3714
3715
3716
3717
3718
3719
3720
3721
3722
3723
3724
3725
3726
3727
3728
3729
3730
3731
3732
3733
3734
3735
3736
3737
3738
3739
3740
3741
3742
3743
3744
3745
3746
3747
3748
3749
3750
3751
3752
3753
3754
3755
3756
3757
3758
3759
3760
3761
3762
3763
3764
3765
3766
3767
3768
3769
3770
3771
3772
3773
3774
3775
3776
3777
3778
3779
3780
3781
3782
3783
3784
3785
3786
3787
3788
3789
3790
3791
3792
3793
3794
3795
3796
3797
3798
3799
3800
3801
3802
3803
3804
3805
3806
3807
3808
3809
3810
3811
3812
3813
3814
3815
3816
3817
3818
3819
3820
3821
3822
3823
3824
3825
3826
3827
3828
3829
3830
3831
3832
3833
3834
3835
3836
3837
3838
3839
3840
3841
3842
3843
3844
3845
3846
3847
3848
3849
3850
3851
3852
3853
3854
3855
3856
3857
3858
3859
3860
3861
3862
3863
3864
3865
3866
3867
3868
3869
3870
3871
3872
3873
3874
3875
3876
3877
3878
3879
3880
3881
3882
3883
3884
3885
3886
3887
3888
3889
3890
3891
3892
3893
3894
3895
3896
3897
3898
3899
3900
3901
3902
3903
3904
3905
3906
3907
3908
3909
3910
3911
3912
3913
3914
3915
3916
3917
3918
3919
3920
3921
3922
3923
3924
3925
3926
3927
3928
3929
3930
3931
3932
3933
3934
3935
3936
3937
3938
3939
3940
3941
3942
3943
3944
3945
3946
3947
3948
3949
3950
3951
3952
3953
3954
3955
3956
3957
3958
3959
3960
3961
3962
3963
3964
3965
3966
3967
3968
3969
3970
3971
3972
3973
3974
3975
3976
3977
3978
3979
3980
3981
3982
3983
3984
3985
3986
3987
3988
3989
3990
3991
3992
3993
3994
3995
3996
3997
3998
3999
4000
4001
4002
4003
4004
4005
4006
4007
4008
4009
4010
4011
4012
4013
4014
4015
4016
4017
4018
4019
4020
4021
4022
4023
4024
4025
4026
4027
4028
4029
4030
4031
4032
4033
4034
4035
4036
4037
4038
4039
4040
4041
4042
4043
4044
4045
4046
4047
4048
4049
4050
4051
4052
4053
4054
4055
4056
4057
4058
4059
4060
4061
4062
4063
4064
4065
4066
4067
4068
4069
4070
4071
4072
4073
4074
4075
4076
4077
4078
4079
4080
4081
4082
4083
4084
4085
4086
4087
4088
4089
4090
4091
4092
4093
4094
4095
4096
4097
4098
4099
4100
4101
4102
4103
4104
4105
4106
4107
4108
4109
4110
4111
4112
4113
4114
4115
4116
4117
4118
4119
4120
4121
4122
4123
4124
4125
4126
4127
4128
4129
4130
4131
4132
4133
4134
4135
4136
4137
4138
4139
4140
4141
4142
4143
4144
4145
4146
4147
4148
4149
4150
4151
4152
4153
4154
4155
4156
4157
4158
4159
4160
4161
4162
4163
4164
4165
4166
4167
4168
4169
4170
4171
4172
4173
4174
4175
4176
4177
4178
4179
4180
4181
4182
4183
4184
4185
4186
4187
4188
4189
4190
4191
4192
4193
4194
4195
4196
4197
4198
4199
4200
4201
4202
4203
4204
4205
4206
4207
4208
4209
4210
4211
4212
4213
4214
4215
4216
4217
4218
4219
4220
4221
4222
4223
4224
4225
4226
4227
4228
4229
4230
4231
4232
4233
4234
4235
4236
4237
4238
4239
4240
4241
4242
4243
4244
4245
4246
4247
4248
4249
4250
4251
4252
4253
4254
4255
4256
4257
4258
4259
4260
4261
4262
4263
4264
4265
4266
4267
4268
4269
4270
4271
4272
4273
4274
4275
4276
4277
4278
4279
4280
4281
4282
4283
4284
4285
4286
4287
4288
4289
4290
4291
4292
4293
4294
4295
4296
4297
4298
4299
4300
4301
4302
4303
4304
4305
4306
4307
4308
4309
4310
4311
4312
4313
4314
4315
4316
4317
4318
4319
4320
4321
4322
4323
4324
4325
4326
4327
4328
4329
4330
4331
4332
4333
4334
4335
4336
4337
4338
4339
4340
4341
4342
4343
4344
4345
4346
4347
4348
4349
4350
4351
4352
4353
4354
4355
4356
4357
4358
4359
4360
4361
4362
4363
4364
4365
4366
4367
4368
4369
4370
4371
4372
4373
4374
4375
4376
4377
4378
4379
4380
4381
4382
4383
4384
4385
4386
4387
4388
4389
4390
4391
4392
4393
4394
4395
4396
4397
4398
4399
4400
4401
4402
4403
4404
4405
4406
4407
4408
4409
4410
4411
4412
4413
4414
4415
4416
4417
4418
4419
4420
4421
4422
4423
4424
4425
4426
4427
4428
4429
4430
4431
4432
4433
4434
4435
4436
4437
4438
4439
4440
4441
4442
4443
4444
4445
4446
4447
4448
4449
4450
4451
4452
4453
4454
4455
4456
4457
4458
4459
4460
4461
4462
4463
4464
4465
4466
4467
4468
4469
4470
4471
4472
4473
4474
4475
4476
4477
4478
4479
4480
4481
4482
4483
4484
4485
4486
4487
4488
4489
4490
4491
4492
4493
4494
4495
4496
4497
4498
4499
4500
4501
4502
4503
4504
4505
4506
4507
4508
4509
4510
4511
4512
4513
4514
4515
4516
4517
4518
4519
4520
4521
4522
4523
4524
4525
4526
4527
4528
4529
4530
4531
4532
4533
4534
4535
4536
4537
4538
4539
4540
4541
4542
4543
4544
4545
4546
4547
4548
4549
4550
4551
4552
4553
4554
4555
4556
4557
4558
4559
4560
4561
4562
4563
4564
4565
4566
4567
4568
4569
4570
4571
4572
4573
4574
4575
4576
4577
4578
4579
4580
4581
4582
4583
4584
4585
4586
4587
4588
4589
4590
4591
4592
4593
4594
4595
4596
4597
4598
4599
4600
4601
4602
4603
4604
4605
4606
4607
4608
4609
4610
4611
4612
4613
4614
4615
4616
4617
4618
4619
4620
4621
4622
4623
4624
4625
4626
4627
4628
4629
4630
4631
4632
4633
4634
4635
4636
4637
4638
4639
4640
4641
4642
4643
4644
4645
4646
4647
4648
4649
4650
4651
4652
4653
4654
4655
4656
4657
4658
4659
4660
4661
4662
4663
4664
4665
4666
4667
4668
4669
4670
4671
4672
4673
4674
4675
4676
4677
4678
4679
4680
4681
4682
4683
4684
4685
4686
4687
4688
4689
4690
4691
4692
4693
4694
4695
4696
4697
4698
4699
4700
4701
4702
4703
4704
4705
4706
4707
4708
4709
4710
4711
4712
4713
4714
4715
4716
4717
4718
4719
4720
4721
4722
4723
4724
4725
4726
4727
4728
4729
4730
4731
4732
4733
4734
4735
4736
4737
4738
4739
4740
4741
4742
4743
4744
4745
4746
4747
4748
4749
4750
4751
4752
4753
4754
4755
4756
4757
4758
4759
4760
4761
4762
4763
4764
4765
4766
4767
4768
4769
4770
4771
4772
4773
4774
4775
4776
4777
4778
4779
4780
4781
4782
4783
4784
4785
4786
4787
4788
4789
4790
4791
4792
4793
4794
4795
4796
4797
4798
4799
4800
4801
4802
4803
4804
4805
4806
4807
4808
4809
4810
4811
4812
4813
4814
4815
4816
4817
4818
4819
4820
4821
4822
4823
4824
4825
4826
4827
4828
4829
4830
4831
4832
4833
4834
4835
4836
4837
4838
4839
4840
4841
4842
4843
4844
4845
4846
4847
4848
4849
4850
4851
4852
4853
4854
4855
4856
4857
4858
4859
4860
4861
4862
4863
4864
4865
4866
4867
4868
4869
4870
4871
4872
4873
4874
4875
4876
4877
4878
4879
4880
4881
4882
4883
4884
4885
4886
4887
4888
4889
4890
4891
4892
4893
4894
4895
4896
4897
4898
4899
4900
4901
4902
4903
4904
4905
4906
4907
4908
4909
4910
4911
4912
4913
4914
4915
4916
4917
4918
4919
4920
4921
4922
4923
4924
4925
4926
4927
4928
4929
4930
4931
4932
4933
4934
4935
4936
4937
4938
4939
4940
4941
4942
4943
4944
4945
4946
4947
4948
4949
4950
4951
4952
4953
4954
4955
4956
4957
4958
4959
4960
4961
4962
4963
4964
4965
4966
4967
4968
4969
4970
4971
4972
4973
4974
4975
4976
4977
4978
4979
4980
4981
4982
4983
4984
4985
4986
4987
4988
4989
4990
4991
4992
4993
4994
4995
4996
4997
4998
4999
5000
5001
5002
5003
5004
5005
5006
5007
5008
5009
5010
5011
5012
5013
5014
5015
5016
5017
5018
5019
5020
5021
5022
5023
5024
5025
5026
5027
5028
5029
5030
5031
5032
5033
5034
5035
5036
5037
5038
5039
5040
5041
5042
5043
5044
5045
5046
5047
5048
5049
5050
5051
5052
5053
5054
5055
5056
5057
5058
5059
5060
5061
5062
5063
5064
5065
5066
5067
5068
5069
5070
5071
5072
5073
5074
5075
5076
5077
5078
5079
5080
5081
5082
5083
5084
5085
5086
5087
5088
5089
5090
5091
5092
5093
5094
5095
5096
5097
5098
5099
5100
5101
5102
5103
5104
5105
5106
5107
5108
5109
5110
5111
5112
5113
5114
5115
5116
5117
5118
5119
5120
5121
5122
5123
5124
5125
5126
5127
5128
5129
5130
5131
5132
5133
5134
5135
5136
5137
5138
5139
5140
5141
5142
5143
5144
5145
5146
5147
5148
5149
5150
5151
5152
5153
5154
5155
5156
5157
5158
5159
5160
5161
5162
5163
5164
5165
5166
5167
5168
5169
5170
5171
5172
5173
5174
5175
5176
5177
5178
5179
5180
5181
5182
5183
5184
5185
5186
5187
5188
5189
5190
5191
5192
5193
5194
5195
5196
5197
5198
5199
5200
5201
5202
5203
5204
5205
5206
5207
5208
5209
5210
5211
5212
5213
5214
5215
5216
5217
5218
5219
5220
5221
5222
5223
5224
5225
5226
5227
5228
5229
5230
5231
5232
5233
5234
5235
5236
5237
5238
5239
5240
5241
5242
5243
5244
5245
5246
5247
5248
5249
5250
5251
5252
5253
5254
5255
5256
5257
5258
5259
5260
5261
5262
5263
5264
5265
5266
5267
5268
5269
5270
5271
5272
5273
5274
5275
5276
5277
5278
5279
5280
5281
5282
5283
5284
5285
5286
5287
5288
5289
5290
5291
5292
5293
5294
5295
5296
5297
5298
5299
5300
5301
5302
5303
5304
5305
5306
5307
5308
5309
5310
5311
5312
5313
5314
5315
5316
5317
5318
5319
5320
5321
5322
5323
5324
5325
5326
5327
5328
5329
5330
5331
5332
5333
5334
5335
5336
5337
5338
5339
5340
5341
5342
5343
5344
5345
5346
5347
5348
5349
5350
5351
5352
5353
5354
5355
5356
5357
5358
5359
5360
5361
5362
5363
5364
5365
5366
5367
5368
5369
5370
5371
5372
5373
5374
5375
5376
5377
5378
5379
5380
5381
5382
5383
5384
5385
5386
5387
5388
5389
5390
5391
5392
5393
5394
5395
5396
5397
5398
5399
5400
5401
5402
5403
5404
5405
5406
5407
5408
5409
5410
5411
5412
5413
5414
5415
5416
5417
5418
5419
5420
5421
5422
5423
5424
5425
5426
5427
5428
5429
5430
5431
5432
5433
5434
5435
5436
5437
5438
5439
5440
5441
5442
5443
5444
5445
5446
5447
5448
5449
5450
5451
5452
5453
5454
5455
5456
5457
5458
5459
5460
5461
5462
5463
5464
5465
5466
5467
5468
5469
5470
5471
5472
5473
5474
5475
5476
5477
5478
5479
5480
5481
5482
5483
5484
5485
5486
5487
5488
5489
5490
5491
5492
5493
5494
5495
5496
5497
5498
5499
5500
5501
5502
5503
5504
5505
5506
5507
5508
5509
5510
5511
5512
5513
5514
5515
5516
5517
5518
5519
5520
5521
5522
5523
5524
5525
5526
5527
5528
5529
5530
5531
5532
5533
5534
5535
5536
5537
5538
5539
5540
5541
5542
5543
5544
5545
5546
5547
5548
5549
5550
5551
5552
5553
5554
5555
5556
5557
5558
5559
5560
5561
5562
5563
5564
5565
5566
5567
5568
5569
5570
5571
5572
5573
5574
5575
5576
5577
5578
5579
5580
5581
5582
5583
5584
5585
5586
5587
5588
5589
5590
5591
5592
5593
5594
5595
5596
5597
5598
5599
5600
5601
5602
5603
5604
5605
5606
5607
5608
5609
5610
5611
5612
5613
5614
5615
5616
5617
5618
5619
5620
5621
5622
5623
5624
5625
5626
5627
5628
5629
5630
5631
5632
5633
5634
5635
5636
5637
5638
5639
5640
5641
5642
5643
5644
5645
5646
5647
5648
5649
5650
5651
5652
5653
5654
5655
5656
5657
5658
5659
5660
5661
5662
5663
5664
5665
5666
5667
5668
5669
5670
5671
5672
5673
5674
5675
5676
5677
5678
5679
5680
5681
5682
5683
5684
5685
5686
5687
5688
5689
5690
5691
5692
5693
5694
5695
5696
5697
5698
5699
5700
5701
5702
5703
5704
5705
5706
5707
5708
5709
5710
5711
5712
5713
5714
5715
5716
5717
5718
5719
5720
5721
5722
5723
5724
5725
5726
5727
5728
5729
5730
5731
5732
5733
5734
5735
5736
5737
5738
5739
5740
5741
5742
5743
5744
5745
5746
5747
5748
5749
5750
5751
5752
5753
5754
5755
5756
5757
5758
5759
5760
5761
5762
5763
5764
5765
5766
5767
5768
5769
5770
5771
5772
5773
5774
5775
5776
5777
5778
5779
5780
5781
5782
5783
5784
5785
5786
5787
5788
5789
5790
5791
5792
5793
5794
5795
5796
5797
5798
5799
5800
5801
5802
5803
5804
5805
5806
5807
5808
5809
5810
5811
5812
5813
5814
5815
5816
5817
5818
5819
5820
5821
5822
5823
5824
5825
5826
5827
5828
5829
5830
5831
5832
5833
5834
5835
5836
5837
5838
5839
5840
5841
5842
5843
5844
5845
5846
5847
5848
5849
5850
5851
5852
5853
5854
5855
5856
5857
5858
5859
5860
5861
5862
5863
5864
5865
5866
5867
5868
5869
5870
5871
5872
5873
5874
5875
5876
5877
5878
5879
5880
5881
5882
5883
5884
5885
5886
5887
5888
5889
5890
5891
5892
5893
5894
5895
5896
5897
5898
5899
5900
5901
5902
5903
5904
5905
5906
5907
5908
5909
5910
5911
5912
5913
|
// -*- mode:C++; tab-width:8; c-basic-offset:2; indent-tabs-mode:t -*-
// vim: ts=8 sw=2 smarttab
/*
* Ceph - scalable distributed file system
*
* Copyright (C) 2004-2006 Sage Weil <sage@newdream.net>
* Copyright (C) 2013,2014 Cloudwatt <libre.licensing@cloudwatt.com>
*
* Author: Loic Dachary <loic@dachary.org>
*
* This is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License version 2.1, as published by the Free Software
* Foundation. See file COPYING.
*
*/
#ifndef CEPH_OSD_TYPES_H
#define CEPH_OSD_TYPES_H
#include <sstream>
#include <stdio.h>
#include <memory>
#include <string_view>
#include <boost/scoped_ptr.hpp>
#include <boost/optional/optional_io.hpp>
#include <boost/variant.hpp>
#include "include/rados/rados_types.hpp"
#include "include/mempool.h"
#include "msg/msg_types.h"
#include "include/types.h"
#include "include/utime.h"
#include "include/CompatSet.h"
#include "common/histogram.h"
#include "include/interval_set.h"
#include "include/inline_memory.h"
#include "common/Formatter.h"
#include "common/bloom_filter.hpp"
#include "common/hobject.h"
#include "common/snap_types.h"
#include "HitSet.h"
#include "Watch.h"
#include "include/cmp.h"
#include "librados/ListObjectImpl.h"
#include "compressor/Compressor.h"
#include <atomic>
#define CEPH_OSD_ONDISK_MAGIC "ceph osd volume v026"
#define CEPH_OSD_FEATURE_INCOMPAT_BASE CompatSet::Feature(1, "initial feature set(~v.18)")
#define CEPH_OSD_FEATURE_INCOMPAT_PGINFO CompatSet::Feature(2, "pginfo object")
#define CEPH_OSD_FEATURE_INCOMPAT_OLOC CompatSet::Feature(3, "object locator")
#define CEPH_OSD_FEATURE_INCOMPAT_LEC CompatSet::Feature(4, "last_epoch_clean")
#define CEPH_OSD_FEATURE_INCOMPAT_CATEGORIES CompatSet::Feature(5, "categories")
#define CEPH_OSD_FEATURE_INCOMPAT_HOBJECTPOOL CompatSet::Feature(6, "hobjectpool")
#define CEPH_OSD_FEATURE_INCOMPAT_BIGINFO CompatSet::Feature(7, "biginfo")
#define CEPH_OSD_FEATURE_INCOMPAT_LEVELDBINFO CompatSet::Feature(8, "leveldbinfo")
#define CEPH_OSD_FEATURE_INCOMPAT_LEVELDBLOG CompatSet::Feature(9, "leveldblog")
#define CEPH_OSD_FEATURE_INCOMPAT_SNAPMAPPER CompatSet::Feature(10, "snapmapper")
#define CEPH_OSD_FEATURE_INCOMPAT_SHARDS CompatSet::Feature(11, "sharded objects")
#define CEPH_OSD_FEATURE_INCOMPAT_HINTS CompatSet::Feature(12, "transaction hints")
#define CEPH_OSD_FEATURE_INCOMPAT_PGMETA CompatSet::Feature(13, "pg meta object")
#define CEPH_OSD_FEATURE_INCOMPAT_MISSING CompatSet::Feature(14, "explicit missing set")
#define CEPH_OSD_FEATURE_INCOMPAT_FASTINFO CompatSet::Feature(15, "fastinfo pg attr")
#define CEPH_OSD_FEATURE_INCOMPAT_RECOVERY_DELETES CompatSet::Feature(16, "deletes in missing set")
/// pool priority range set by user
#define OSD_POOL_PRIORITY_MAX 10
#define OSD_POOL_PRIORITY_MIN -OSD_POOL_PRIORITY_MAX
/// min recovery priority for MBackfillReserve
#define OSD_RECOVERY_PRIORITY_MIN 0
/// base backfill priority for MBackfillReserve
#define OSD_BACKFILL_PRIORITY_BASE 100
/// base backfill priority for MBackfillReserve (degraded PG)
#define OSD_BACKFILL_DEGRADED_PRIORITY_BASE 140
/// base recovery priority for MBackfillReserve
#define OSD_RECOVERY_PRIORITY_BASE 180
/// base backfill priority for MBackfillReserve (inactive PG)
#define OSD_BACKFILL_INACTIVE_PRIORITY_BASE 220
/// base recovery priority for MRecoveryReserve (inactive PG)
#define OSD_RECOVERY_INACTIVE_PRIORITY_BASE 220
/// max manually/automatically set recovery priority for MBackfillReserve
#define OSD_RECOVERY_PRIORITY_MAX 253
/// backfill priority for MBackfillReserve, when forced manually
#define OSD_BACKFILL_PRIORITY_FORCED 254
/// recovery priority for MRecoveryReserve, when forced manually
#define OSD_RECOVERY_PRIORITY_FORCED 255
/// priority for pg deletion when osd is not fullish
#define OSD_DELETE_PRIORITY_NORMAL 179
/// priority for pg deletion when osd is approaching full
#define OSD_DELETE_PRIORITY_FULLISH 219
/// priority when more full
#define OSD_DELETE_PRIORITY_FULL 255
static std::map<int, int> max_prio_map = {
{OSD_BACKFILL_PRIORITY_BASE, OSD_BACKFILL_DEGRADED_PRIORITY_BASE - 1},
{OSD_BACKFILL_DEGRADED_PRIORITY_BASE, OSD_RECOVERY_PRIORITY_BASE - 1},
{OSD_RECOVERY_PRIORITY_BASE, OSD_BACKFILL_INACTIVE_PRIORITY_BASE - 1},
{OSD_RECOVERY_INACTIVE_PRIORITY_BASE, OSD_RECOVERY_PRIORITY_MAX},
{OSD_BACKFILL_INACTIVE_PRIORITY_BASE, OSD_RECOVERY_PRIORITY_MAX}
};
typedef hobject_t collection_list_handle_t;
/// convert a single CPEH_OSD_FLAG_* to a string
const char *ceph_osd_flag_name(unsigned flag);
/// convert a single CEPH_OSD_OF_FLAG_* to a string
const char *ceph_osd_op_flag_name(unsigned flag);
/// convert CEPH_OSD_FLAG_* op flags to a string
string ceph_osd_flag_string(unsigned flags);
/// conver CEPH_OSD_OP_FLAG_* op flags to a string
string ceph_osd_op_flag_string(unsigned flags);
/// conver CEPH_OSD_ALLOC_HINT_FLAG_* op flags to a string
string ceph_osd_alloc_hint_flag_string(unsigned flags);
typedef map<string,string> osd_alert_list_t;
/// map osd id -> alert_list_t
typedef map<int, osd_alert_list_t> osd_alerts_t;
void dump(Formatter* f, const osd_alerts_t& alerts);
/**
* osd request identifier
*
* caller name + incarnation# + tid to unique identify this request.
*/
struct osd_reqid_t {
entity_name_t name; // who
ceph_tid_t tid;
int32_t inc; // incarnation
osd_reqid_t()
: tid(0), inc(0)
{}
osd_reqid_t(const osd_reqid_t& other)
: name(other.name), tid(other.tid), inc(other.inc)
{}
osd_reqid_t(const entity_name_t& a, int i, ceph_tid_t t)
: name(a), tid(t), inc(i)
{}
DENC(osd_reqid_t, v, p) {
DENC_START(2, 2, p);
denc(v.name, p);
denc(v.tid, p);
denc(v.inc, p);
DENC_FINISH(p);
}
void dump(Formatter *f) const;
static void generate_test_instances(list<osd_reqid_t*>& o);
};
WRITE_CLASS_DENC(osd_reqid_t)
struct pg_shard_t {
static const int32_t NO_OSD = 0x7fffffff;
int32_t osd;
shard_id_t shard;
pg_shard_t() : osd(-1), shard(shard_id_t::NO_SHARD) {}
explicit pg_shard_t(int osd) : osd(osd), shard(shard_id_t::NO_SHARD) {}
pg_shard_t(int osd, shard_id_t shard) : osd(osd), shard(shard) {}
bool is_undefined() const {
return osd == -1;
}
string get_osd() const { return (osd == NO_OSD ? "NONE" : to_string(osd)); }
void encode(bufferlist &bl) const;
void decode(bufferlist::const_iterator &bl);
void dump(Formatter *f) const {
f->dump_unsigned("osd", osd);
if (shard != shard_id_t::NO_SHARD) {
f->dump_unsigned("shard", shard);
}
}
};
WRITE_CLASS_ENCODER(pg_shard_t)
WRITE_EQ_OPERATORS_2(pg_shard_t, osd, shard)
WRITE_CMP_OPERATORS_2(pg_shard_t, osd, shard)
ostream &operator<<(ostream &lhs, const pg_shard_t &rhs);
class IsPGRecoverablePredicate {
public:
/**
* have encodes the shards available
*/
virtual bool operator()(const set<pg_shard_t> &have) const = 0;
virtual ~IsPGRecoverablePredicate() {}
};
class IsPGReadablePredicate {
public:
/**
* have encodes the shards available
*/
virtual bool operator()(const set<pg_shard_t> &have) const = 0;
virtual ~IsPGReadablePredicate() {}
};
inline ostream& operator<<(ostream& out, const osd_reqid_t& r) {
return out << r.name << "." << r.inc << ":" << r.tid;
}
inline bool operator==(const osd_reqid_t& l, const osd_reqid_t& r) {
return (l.name == r.name) && (l.inc == r.inc) && (l.tid == r.tid);
}
inline bool operator!=(const osd_reqid_t& l, const osd_reqid_t& r) {
return (l.name != r.name) || (l.inc != r.inc) || (l.tid != r.tid);
}
inline bool operator<(const osd_reqid_t& l, const osd_reqid_t& r) {
return (l.name < r.name) || (l.inc < r.inc) ||
(l.name == r.name && l.inc == r.inc && l.tid < r.tid);
}
inline bool operator<=(const osd_reqid_t& l, const osd_reqid_t& r) {
return (l.name < r.name) || (l.inc < r.inc) ||
(l.name == r.name && l.inc == r.inc && l.tid <= r.tid);
}
inline bool operator>(const osd_reqid_t& l, const osd_reqid_t& r) { return !(l <= r); }
inline bool operator>=(const osd_reqid_t& l, const osd_reqid_t& r) { return !(l < r); }
namespace std {
template<> struct hash<osd_reqid_t> {
size_t operator()(const osd_reqid_t &r) const {
static hash<uint64_t> H;
return H(r.name.num() ^ r.tid ^ r.inc);
}
};
} // namespace std
// -----
// a locator constrains the placement of an object. mainly, which pool
// does it go in.
struct object_locator_t {
// You specify either the hash or the key -- not both
int64_t pool; ///< pool id
string key; ///< key string (if non-empty)
string nspace; ///< namespace
int64_t hash; ///< hash position (if >= 0)
explicit object_locator_t()
: pool(-1), hash(-1) {}
explicit object_locator_t(int64_t po)
: pool(po), hash(-1) {}
explicit object_locator_t(int64_t po, int64_t ps)
: pool(po), hash(ps) {}
explicit object_locator_t(int64_t po, string ns)
: pool(po), nspace(ns), hash(-1) {}
explicit object_locator_t(int64_t po, string ns, int64_t ps)
: pool(po), nspace(ns), hash(ps) {}
explicit object_locator_t(int64_t po, string ns, string s)
: pool(po), key(s), nspace(ns), hash(-1) {}
explicit object_locator_t(const hobject_t& soid)
: pool(soid.pool), key(soid.get_key()), nspace(soid.nspace), hash(-1) {}
int64_t get_pool() const {
return pool;
}
void clear() {
pool = -1;
key = "";
nspace = "";
hash = -1;
}
bool empty() const {
return pool == -1;
}
void encode(bufferlist& bl) const;
void decode(bufferlist::const_iterator& p);
void dump(Formatter *f) const;
static void generate_test_instances(list<object_locator_t*>& o);
};
WRITE_CLASS_ENCODER(object_locator_t)
inline bool operator==(const object_locator_t& l, const object_locator_t& r) {
return l.pool == r.pool && l.key == r.key && l.nspace == r.nspace && l.hash == r.hash;
}
inline bool operator!=(const object_locator_t& l, const object_locator_t& r) {
return !(l == r);
}
inline ostream& operator<<(ostream& out, const object_locator_t& loc)
{
out << "@" << loc.pool;
if (loc.nspace.length())
out << ";" << loc.nspace;
if (loc.key.length())
out << ":" << loc.key;
return out;
}
struct request_redirect_t {
private:
object_locator_t redirect_locator; ///< this is authoritative
string redirect_object; ///< If non-empty, the request goes to this object name
friend ostream& operator<<(ostream& out, const request_redirect_t& redir);
public:
request_redirect_t() {}
explicit request_redirect_t(const object_locator_t& orig, int64_t rpool) :
redirect_locator(orig) { redirect_locator.pool = rpool; }
explicit request_redirect_t(const object_locator_t& rloc) :
redirect_locator(rloc) {}
explicit request_redirect_t(const object_locator_t& orig,
const string& robj) :
redirect_locator(orig), redirect_object(robj) {}
bool empty() const { return redirect_locator.empty() &&
redirect_object.empty(); }
void combine_with_locator(object_locator_t& orig, string& obj) const {
orig = redirect_locator;
if (!redirect_object.empty())
obj = redirect_object;
}
void encode(bufferlist& bl) const;
void decode(bufferlist::const_iterator& bl);
void dump(Formatter *f) const;
static void generate_test_instances(list<request_redirect_t*>& o);
};
WRITE_CLASS_ENCODER(request_redirect_t)
inline ostream& operator<<(ostream& out, const request_redirect_t& redir) {
out << "object " << redir.redirect_object << ", locator{" << redir.redirect_locator << "}";
return out;
}
// Internal OSD op flags - set by the OSD based on the op types
enum {
CEPH_OSD_RMW_FLAG_READ = (1 << 1),
CEPH_OSD_RMW_FLAG_WRITE = (1 << 2),
CEPH_OSD_RMW_FLAG_CLASS_READ = (1 << 3),
CEPH_OSD_RMW_FLAG_CLASS_WRITE = (1 << 4),
CEPH_OSD_RMW_FLAG_PGOP = (1 << 5),
CEPH_OSD_RMW_FLAG_CACHE = (1 << 6),
CEPH_OSD_RMW_FLAG_FORCE_PROMOTE = (1 << 7),
CEPH_OSD_RMW_FLAG_SKIP_HANDLE_CACHE = (1 << 8),
CEPH_OSD_RMW_FLAG_SKIP_PROMOTE = (1 << 9),
CEPH_OSD_RMW_FLAG_RWORDERED = (1 << 10),
};
// pg stuff
#define OSD_SUPERBLOCK_GOBJECT ghobject_t(hobject_t(sobject_t(object_t("osd_superblock"), 0)))
// placement seed (a hash value)
typedef uint32_t ps_t;
// old (v1) pg_t encoding (wrap old struct ceph_pg)
struct old_pg_t {
ceph_pg v;
void encode(bufferlist& bl) const {
::encode_raw(v, bl);
}
void decode(bufferlist::const_iterator& bl) {
::decode_raw(v, bl);
}
};
WRITE_CLASS_ENCODER(old_pg_t)
// placement group id
struct pg_t {
uint64_t m_pool;
uint32_t m_seed;
pg_t() : m_pool(0), m_seed(0) {}
pg_t(ps_t seed, uint64_t pool) :
m_pool(pool), m_seed(seed) {}
// cppcheck-suppress noExplicitConstructor
pg_t(const ceph_pg& cpg) :
m_pool(cpg.pool), m_seed(cpg.ps) {}
// cppcheck-suppress noExplicitConstructor
pg_t(const old_pg_t& opg) {
*this = opg.v;
}
old_pg_t get_old_pg() const {
old_pg_t o;
ceph_assert(m_pool < 0xffffffffull);
o.v.pool = m_pool;
o.v.ps = m_seed;
o.v.preferred = (__s16)-1;
return o;
}
ps_t ps() const {
return m_seed;
}
int64_t pool() const {
return m_pool;
}
static const uint8_t calc_name_buf_size = 36; // max length for max values len("18446744073709551615.ffffffff") + future suffix len("_head") + '\0'
char *calc_name(char *buf, const char *suffix_backwords) const;
void set_ps(ps_t p) {
m_seed = p;
}
void set_pool(uint64_t p) {
m_pool = p;
}
pg_t get_parent() const;
pg_t get_ancestor(unsigned old_pg_num) const;
int print(char *o, int maxlen) const;
bool parse(const char *s);
bool is_split(unsigned old_pg_num, unsigned new_pg_num, set<pg_t> *pchildren) const;
bool is_merge_source(unsigned old_pg_num, unsigned new_pg_num, pg_t *parent) const;
bool is_merge_target(unsigned old_pg_num, unsigned new_pg_num) const {
return ps() < new_pg_num && is_split(new_pg_num, old_pg_num, nullptr);
}
/**
* Returns b such that for all object o:
* ~((~0)<<b) & o.hash) == 0 iff o is in the pg for *this
*/
unsigned get_split_bits(unsigned pg_num) const;
bool contains(int bits, const ghobject_t& oid) {
return
(int64_t)m_pool == oid.hobj.get_logical_pool() &&
oid.match(bits, ps());
}
bool contains(int bits, const hobject_t& oid) {
return
(int64_t)m_pool == oid.get_logical_pool() &&
oid.match(bits, ps());
}
hobject_t get_hobj_start() const;
hobject_t get_hobj_end(unsigned pg_num) const;
void encode(bufferlist& bl) const {
using ceph::encode;
__u8 v = 1;
encode(v, bl);
encode(m_pool, bl);
encode(m_seed, bl);
encode((int32_t)-1, bl); // was preferred
}
void decode(bufferlist::const_iterator& bl) {
using ceph::decode;
__u8 v;
decode(v, bl);
decode(m_pool, bl);
decode(m_seed, bl);
bl.advance(sizeof(int32_t)); // was preferred
}
void decode_old(bufferlist::const_iterator& bl) {
using ceph::decode;
old_pg_t opg;
decode(opg, bl);
*this = opg;
}
void dump(Formatter *f) const;
static void generate_test_instances(list<pg_t*>& o);
};
WRITE_CLASS_ENCODER(pg_t)
inline bool operator<(const pg_t& l, const pg_t& r) {
return l.pool() < r.pool() ||
(l.pool() == r.pool() && (l.ps() < r.ps()));
}
inline bool operator<=(const pg_t& l, const pg_t& r) {
return l.pool() < r.pool() ||
(l.pool() == r.pool() && (l.ps() <= r.ps()));
}
inline bool operator==(const pg_t& l, const pg_t& r) {
return l.pool() == r.pool() &&
l.ps() == r.ps();
}
inline bool operator!=(const pg_t& l, const pg_t& r) {
return l.pool() != r.pool() ||
l.ps() != r.ps();
}
inline bool operator>(const pg_t& l, const pg_t& r) {
return l.pool() > r.pool() ||
(l.pool() == r.pool() && (l.ps() > r.ps()));
}
inline bool operator>=(const pg_t& l, const pg_t& r) {
return l.pool() > r.pool() ||
(l.pool() == r.pool() && (l.ps() >= r.ps()));
}
ostream& operator<<(ostream& out, const pg_t &pg);
namespace std {
template<> struct hash< pg_t >
{
size_t operator()( const pg_t& x ) const
{
static hash<uint32_t> H;
// xor (s32)-1 in there to preserve original m_preferred result (paranoia!)
return H((x.pool() & 0xffffffff) ^ (x.pool() >> 32) ^ x.ps() ^ (int32_t)(-1));
}
};
} // namespace std
struct spg_t {
pg_t pgid;
shard_id_t shard;
spg_t() : shard(shard_id_t::NO_SHARD) {}
spg_t(pg_t pgid, shard_id_t shard) : pgid(pgid), shard(shard) {}
explicit spg_t(pg_t pgid) : pgid(pgid), shard(shard_id_t::NO_SHARD) {}
unsigned get_split_bits(unsigned pg_num) const {
return pgid.get_split_bits(pg_num);
}
spg_t get_parent() const {
return spg_t(pgid.get_parent(), shard);
}
ps_t ps() const {
return pgid.ps();
}
uint64_t pool() const {
return pgid.pool();
}
static const uint8_t calc_name_buf_size = pg_t::calc_name_buf_size + 4; // 36 + len('s') + len("255");
char *calc_name(char *buf, const char *suffix_backwords) const;
bool parse(const char *s);
bool parse(const std::string& s) {
return parse(s.c_str());
}
spg_t get_ancestor(unsigned old_pg_num) const {
return spg_t(pgid.get_ancestor(old_pg_num), shard);
}
bool is_split(unsigned old_pg_num, unsigned new_pg_num,
set<spg_t> *pchildren) const {
set<pg_t> _children;
set<pg_t> *children = pchildren ? &_children : NULL;
bool is_split = pgid.is_split(old_pg_num, new_pg_num, children);
if (pchildren && is_split) {
for (set<pg_t>::iterator i = _children.begin();
i != _children.end();
++i) {
pchildren->insert(spg_t(*i, shard));
}
}
return is_split;
}
bool is_merge_target(unsigned old_pg_num, unsigned new_pg_num) const {
return pgid.is_merge_target(old_pg_num, new_pg_num);
}
bool is_merge_source(unsigned old_pg_num, unsigned new_pg_num,
spg_t *parent) const {
spg_t out = *this;
bool r = pgid.is_merge_source(old_pg_num, new_pg_num, &out.pgid);
if (r && parent) {
*parent = out;
}
return r;
}
bool is_no_shard() const {
return shard == shard_id_t::NO_SHARD;
}
ghobject_t make_pgmeta_oid() const {
return ghobject_t::make_pgmeta(pgid.pool(), pgid.ps(), shard);
}
void encode(bufferlist &bl) const {
ENCODE_START(1, 1, bl);
encode(pgid, bl);
encode(shard, bl);
ENCODE_FINISH(bl);
}
void decode(bufferlist::const_iterator& bl) {
DECODE_START(1, bl);
decode(pgid, bl);
decode(shard, bl);
DECODE_FINISH(bl);
}
ghobject_t make_temp_ghobject(const string& name) const {
return ghobject_t(
hobject_t(object_t(name), "", CEPH_NOSNAP,
pgid.ps(),
hobject_t::get_temp_pool(pgid.pool()),
""),
ghobject_t::NO_GEN,
shard);
}
unsigned hash_to_shard(unsigned num_shards) const {
return ps() % num_shards;
}
};
WRITE_CLASS_ENCODER(spg_t)
WRITE_EQ_OPERATORS_2(spg_t, pgid, shard)
WRITE_CMP_OPERATORS_2(spg_t, pgid, shard)
namespace std {
template<> struct hash< spg_t >
{
size_t operator()( const spg_t& x ) const
{
static hash<uint32_t> H;
return H(hash<pg_t>()(x.pgid) ^ x.shard);
}
};
} // namespace std
ostream& operator<<(ostream& out, const spg_t &pg);
// ----------------------
class coll_t {
enum type_t {
TYPE_META = 0,
TYPE_LEGACY_TEMP = 1, /* no longer used */
TYPE_PG = 2,
TYPE_PG_TEMP = 3,
};
type_t type;
spg_t pgid;
uint64_t removal_seq; // note: deprecated, not encoded
char _str_buff[spg_t::calc_name_buf_size];
char *_str;
void calc_str();
coll_t(type_t t, spg_t p, uint64_t r)
: type(t), pgid(p), removal_seq(r) {
calc_str();
}
public:
coll_t() : type(TYPE_META), removal_seq(0)
{
calc_str();
}
coll_t(const coll_t& other)
: type(other.type), pgid(other.pgid), removal_seq(other.removal_seq) {
calc_str();
}
explicit coll_t(spg_t pgid)
: type(TYPE_PG), pgid(pgid), removal_seq(0)
{
calc_str();
}
coll_t& operator=(const coll_t& rhs)
{
this->type = rhs.type;
this->pgid = rhs.pgid;
this->removal_seq = rhs.removal_seq;
this->calc_str();
return *this;
}
// named constructors
static coll_t meta() {
return coll_t();
}
static coll_t pg(spg_t p) {
return coll_t(p);
}
const std::string to_str() const {
return string(_str);
}
const char *c_str() const {
return _str;
}
bool parse(const std::string& s);
int operator<(const coll_t &rhs) const {
return type < rhs.type ||
(type == rhs.type && pgid < rhs.pgid);
}
bool is_meta() const {
return type == TYPE_META;
}
bool is_pg_prefix(spg_t *pgid_) const {
if (type == TYPE_PG || type == TYPE_PG_TEMP) {
*pgid_ = pgid;
return true;
}
return false;
}
bool is_pg() const {
return type == TYPE_PG;
}
bool is_pg(spg_t *pgid_) const {
if (type == TYPE_PG) {
*pgid_ = pgid;
return true;
}
return false;
}
bool is_temp() const {
return type == TYPE_PG_TEMP;
}
bool is_temp(spg_t *pgid_) const {
if (type == TYPE_PG_TEMP) {
*pgid_ = pgid;
return true;
}
return false;
}
void encode(bufferlist& bl) const;
void decode(bufferlist::const_iterator& bl);
size_t encoded_size() const;
inline bool operator==(const coll_t& rhs) const {
// only compare type if meta
if (type != rhs.type)
return false;
if (type == TYPE_META)
return true;
return type == rhs.type && pgid == rhs.pgid;
}
inline bool operator!=(const coll_t& rhs) const {
return !(*this == rhs);
}
// get a TEMP collection that corresponds to the current collection,
// which we presume is a pg collection.
coll_t get_temp() const {
ceph_assert(type == TYPE_PG);
return coll_t(TYPE_PG_TEMP, pgid, 0);
}
ghobject_t get_min_hobj() const {
ghobject_t o;
switch (type) {
case TYPE_PG:
o.hobj.pool = pgid.pool();
o.set_shard(pgid.shard);
break;
case TYPE_META:
o.hobj.pool = -1;
break;
default:
break;
}
return o;
}
unsigned hash_to_shard(unsigned num_shards) const {
if (type == TYPE_PG)
return pgid.hash_to_shard(num_shards);
return 0; // whatever.
}
void dump(Formatter *f) const;
static void generate_test_instances(list<coll_t*>& o);
};
WRITE_CLASS_ENCODER(coll_t)
inline ostream& operator<<(ostream& out, const coll_t& c) {
out << c.to_str();
return out;
}
namespace std {
template<> struct hash<coll_t> {
size_t operator()(const coll_t &c) const {
size_t h = 0;
string str(c.to_str());
std::string::const_iterator end(str.end());
for (std::string::const_iterator s = str.begin(); s != end; ++s) {
h += *s;
h += (h << 10);
h ^= (h >> 6);
}
h += (h << 3);
h ^= (h >> 11);
h += (h << 15);
return h;
}
};
} // namespace std
inline ostream& operator<<(ostream& out, const ceph_object_layout &ol)
{
out << pg_t(ol.ol_pgid);
int su = ol.ol_stripe_unit;
if (su)
out << ".su=" << su;
return out;
}
// compound rados version type
/* WARNING: If add member in eversion_t, please make sure the encode/decode function
* work well. For little-endian machine, we should make sure there is no padding
* in 32-bit machine and 64-bit machine.
*/
class eversion_t {
public:
version_t version;
epoch_t epoch;
__u32 __pad;
eversion_t() : version(0), epoch(0), __pad(0) {}
eversion_t(epoch_t e, version_t v) : version(v), epoch(e), __pad(0) {}
// cppcheck-suppress noExplicitConstructor
eversion_t(const ceph_eversion& ce) :
version(ce.version),
epoch(ce.epoch),
__pad(0) { }
explicit eversion_t(bufferlist& bl) : __pad(0) { decode(bl); }
static const eversion_t& max() {
static const eversion_t max(-1,-1);
return max;
}
operator ceph_eversion() {
ceph_eversion c;
c.epoch = epoch;
c.version = version;
return c;
}
string get_key_name() const;
// key must point to the beginning of a block of 32 chars
inline void get_key_name(char* key) const {
// Below is equivalent of sprintf("%010u.%020llu");
key[31] = 0;
ritoa<uint64_t, 10, 20>(version, key + 31);
key[10] = '.';
ritoa<uint32_t, 10, 10>(epoch, key + 10);
}
void encode(bufferlist &bl) const {
#if defined(CEPH_LITTLE_ENDIAN)
bl.append((char *)this, sizeof(version_t) + sizeof(epoch_t));
#else
using ceph::encode;
encode(version, bl);
encode(epoch, bl);
#endif
}
void decode(bufferlist::const_iterator &bl) {
#if defined(CEPH_LITTLE_ENDIAN)
bl.copy(sizeof(version_t) + sizeof(epoch_t), (char *)this);
#else
using ceph::decode;
decode(version, bl);
decode(epoch, bl);
#endif
}
void decode(bufferlist& bl) {
auto p = std::cbegin(bl);
decode(p);
}
};
WRITE_CLASS_ENCODER(eversion_t)
inline bool operator==(const eversion_t& l, const eversion_t& r) {
return (l.epoch == r.epoch) && (l.version == r.version);
}
inline bool operator!=(const eversion_t& l, const eversion_t& r) {
return (l.epoch != r.epoch) || (l.version != r.version);
}
inline bool operator<(const eversion_t& l, const eversion_t& r) {
return (l.epoch == r.epoch) ? (l.version < r.version):(l.epoch < r.epoch);
}
inline bool operator<=(const eversion_t& l, const eversion_t& r) {
return (l.epoch == r.epoch) ? (l.version <= r.version):(l.epoch <= r.epoch);
}
inline bool operator>(const eversion_t& l, const eversion_t& r) {
return (l.epoch == r.epoch) ? (l.version > r.version):(l.epoch > r.epoch);
}
inline bool operator>=(const eversion_t& l, const eversion_t& r) {
return (l.epoch == r.epoch) ? (l.version >= r.version):(l.epoch >= r.epoch);
}
inline ostream& operator<<(ostream& out, const eversion_t& e) {
return out << e.epoch << "'" << e.version;
}
/**
* objectstore_perf_stat_t
*
* current perf information about the osd
*/
struct objectstore_perf_stat_t {
// cur_op_latency is in ns since double add/sub are not associative
uint64_t os_commit_latency_ns;
uint64_t os_apply_latency_ns;
objectstore_perf_stat_t() :
os_commit_latency_ns(0), os_apply_latency_ns(0) {}
bool operator==(const objectstore_perf_stat_t &r) const {
return os_commit_latency_ns == r.os_commit_latency_ns &&
os_apply_latency_ns == r.os_apply_latency_ns;
}
void add(const objectstore_perf_stat_t &o) {
os_commit_latency_ns += o.os_commit_latency_ns;
os_apply_latency_ns += o.os_apply_latency_ns;
}
void sub(const objectstore_perf_stat_t &o) {
os_commit_latency_ns -= o.os_commit_latency_ns;
os_apply_latency_ns -= o.os_apply_latency_ns;
}
void dump(Formatter *f) const;
void encode(bufferlist &bl, uint64_t features) const;
void decode(bufferlist::const_iterator &bl);
static void generate_test_instances(std::list<objectstore_perf_stat_t*>& o);
};
WRITE_CLASS_ENCODER_FEATURES(objectstore_perf_stat_t)
/*
* pg states
*/
#define PG_STATE_CREATING (1ULL << 0) // creating
#define PG_STATE_ACTIVE (1ULL << 1) // i am active. (primary: replicas too)
#define PG_STATE_CLEAN (1ULL << 2) // peers are complete, clean of stray replicas.
#define PG_STATE_DOWN (1ULL << 4) // a needed replica is down, PG offline
#define PG_STATE_RECOVERY_UNFOUND (1ULL << 5) // recovery stopped due to unfound
#define PG_STATE_BACKFILL_UNFOUND (1ULL << 6) // backfill stopped due to unfound
#define PG_STATE_PREMERGE (1ULL << 7) // i am prepare to merging
#define PG_STATE_SCRUBBING (1ULL << 8) // scrubbing
//#define PG_STATE_SCRUBQ (1ULL << 9) // queued for scrub
#define PG_STATE_DEGRADED (1ULL << 10) // pg contains objects with reduced redundancy
#define PG_STATE_INCONSISTENT (1ULL << 11) // pg replicas are inconsistent (but shouldn't be)
#define PG_STATE_PEERING (1ULL << 12) // pg is (re)peering
#define PG_STATE_REPAIR (1ULL << 13) // pg should repair on next scrub
#define PG_STATE_RECOVERING (1ULL << 14) // pg is recovering/migrating objects
#define PG_STATE_BACKFILL_WAIT (1ULL << 15) // [active] reserving backfill
#define PG_STATE_INCOMPLETE (1ULL << 16) // incomplete content, peering failed.
#define PG_STATE_STALE (1ULL << 17) // our state for this pg is stale, unknown.
#define PG_STATE_REMAPPED (1ULL << 18) // pg is explicitly remapped to different OSDs than CRUSH
#define PG_STATE_DEEP_SCRUB (1ULL << 19) // deep scrub: check CRC32 on files
#define PG_STATE_BACKFILLING (1ULL << 20) // [active] backfilling pg content
#define PG_STATE_BACKFILL_TOOFULL (1ULL << 21) // backfill can't proceed: too full
#define PG_STATE_RECOVERY_WAIT (1ULL << 22) // waiting for recovery reservations
#define PG_STATE_UNDERSIZED (1ULL << 23) // pg acting < pool size
#define PG_STATE_ACTIVATING (1ULL << 24) // pg is peered but not yet active
#define PG_STATE_PEERED (1ULL << 25) // peered, cannot go active, can recover
#define PG_STATE_SNAPTRIM (1ULL << 26) // trimming snaps
#define PG_STATE_SNAPTRIM_WAIT (1ULL << 27) // queued to trim snaps
#define PG_STATE_RECOVERY_TOOFULL (1ULL << 28) // recovery can't proceed: too full
#define PG_STATE_SNAPTRIM_ERROR (1ULL << 29) // error stopped trimming snaps
#define PG_STATE_FORCED_RECOVERY (1ULL << 30) // force recovery of this pg before any other
#define PG_STATE_FORCED_BACKFILL (1ULL << 31) // force backfill of this pg before any other
#define PG_STATE_FAILED_REPAIR (1ULL << 32) // A repair failed to fix all errors
std::string pg_state_string(uint64_t state);
std::string pg_vector_string(const vector<int32_t> &a);
boost::optional<uint64_t> pg_string_state(const std::string& state);
/*
* pool_snap_info_t
*
* attributes for a single pool snapshot.
*/
struct pool_snap_info_t {
snapid_t snapid;
utime_t stamp;
string name;
void dump(Formatter *f) const;
void encode(bufferlist& bl, uint64_t features) const;
void decode(bufferlist::const_iterator& bl);
static void generate_test_instances(list<pool_snap_info_t*>& o);
};
WRITE_CLASS_ENCODER_FEATURES(pool_snap_info_t)
inline ostream& operator<<(ostream& out, const pool_snap_info_t& si) {
return out << si.snapid << '(' << si.name << ' ' << si.stamp << ')';
}
/*
* pool_opts_t
*
* pool options.
*/
class pool_opts_t {
public:
enum key_t {
SCRUB_MIN_INTERVAL,
SCRUB_MAX_INTERVAL,
DEEP_SCRUB_INTERVAL,
RECOVERY_PRIORITY,
RECOVERY_OP_PRIORITY,
SCRUB_PRIORITY,
COMPRESSION_MODE,
COMPRESSION_ALGORITHM,
COMPRESSION_REQUIRED_RATIO,
COMPRESSION_MAX_BLOB_SIZE,
COMPRESSION_MIN_BLOB_SIZE,
CSUM_TYPE,
CSUM_MAX_BLOCK,
CSUM_MIN_BLOCK,
FINGERPRINT_ALGORITHM,
PG_NUM_MIN, // min pg_num
TARGET_SIZE_BYTES, // total bytes in pool
TARGET_SIZE_RATIO, // fraction of total cluster
PG_AUTOSCALE_BIAS,
};
enum type_t {
STR,
INT,
DOUBLE,
};
struct opt_desc_t {
key_t key;
type_t type;
opt_desc_t(key_t k, type_t t) : key(k), type(t) {}
bool operator==(const opt_desc_t& rhs) const {
return key == rhs.key && type == rhs.type;
}
};
typedef boost::variant<std::string,int64_t,double> value_t;
static bool is_opt_name(const std::string& name);
static opt_desc_t get_opt_desc(const std::string& name);
pool_opts_t() : opts() {}
bool is_set(key_t key) const;
template<typename T>
void set(key_t key, const T &val) {
value_t value = val;
opts[key] = value;
}
template<typename T>
bool get(key_t key, T *val) const {
opts_t::const_iterator i = opts.find(key);
if (i == opts.end()) {
return false;
}
*val = boost::get<T>(i->second);
return true;
}
const value_t& get(key_t key) const;
bool unset(key_t key);
void dump(const std::string& name, Formatter *f) const;
void dump(Formatter *f) const;
void encode(bufferlist &bl, uint64_t features) const;
void decode(bufferlist::const_iterator &bl);
private:
typedef std::map<key_t, value_t> opts_t;
opts_t opts;
friend ostream& operator<<(ostream& out, const pool_opts_t& opts);
};
WRITE_CLASS_ENCODER_FEATURES(pool_opts_t)
struct pg_merge_meta_t {
pg_t source_pgid;
epoch_t ready_epoch = 0;
epoch_t last_epoch_started = 0;
epoch_t last_epoch_clean = 0;
eversion_t source_version;
eversion_t target_version;
void encode(bufferlist& bl) const {
ENCODE_START(1, 1, bl);
encode(source_pgid, bl);
encode(ready_epoch, bl);
encode(last_epoch_started, bl);
encode(last_epoch_clean, bl);
encode(source_version, bl);
encode(target_version, bl);
ENCODE_FINISH(bl);
}
void decode(bufferlist::const_iterator& p) {
DECODE_START(1, p);
decode(source_pgid, p);
decode(ready_epoch, p);
decode(last_epoch_started, p);
decode(last_epoch_clean, p);
decode(source_version, p);
decode(target_version, p);
DECODE_FINISH(p);
}
void dump(Formatter *f) const {
f->dump_stream("source_pgid") << source_pgid;
f->dump_unsigned("ready_epoch", ready_epoch);
f->dump_unsigned("last_epoch_started", last_epoch_started);
f->dump_unsigned("last_epoch_clean", last_epoch_clean);
f->dump_stream("source_version") << source_version;
f->dump_stream("target_version") << target_version;
}
};
WRITE_CLASS_ENCODER(pg_merge_meta_t)
/*
* pg_pool
*/
struct pg_pool_t {
static const char *APPLICATION_NAME_CEPHFS;
static const char *APPLICATION_NAME_RBD;
static const char *APPLICATION_NAME_RGW;
enum {
TYPE_REPLICATED = 1, // replication
//TYPE_RAID4 = 2, // raid4 (never implemented)
TYPE_ERASURE = 3, // erasure-coded
};
static std::string_view get_type_name(int t) {
switch (t) {
case TYPE_REPLICATED: return "replicated";
//case TYPE_RAID4: return "raid4";
case TYPE_ERASURE: return "erasure";
default: return "???";
}
}
std::string_view get_type_name() const {
return get_type_name(type);
}
enum {
FLAG_HASHPSPOOL = 1<<0, // hash pg seed and pool together (instead of adding)
FLAG_FULL = 1<<1, // pool is full
FLAG_EC_OVERWRITES = 1<<2, // enables overwrites, once enabled, cannot be disabled
FLAG_INCOMPLETE_CLONES = 1<<3, // may have incomplete clones (bc we are/were an overlay)
FLAG_NODELETE = 1<<4, // pool can't be deleted
FLAG_NOPGCHANGE = 1<<5, // pool's pg and pgp num can't be changed
FLAG_NOSIZECHANGE = 1<<6, // pool's size and min size can't be changed
FLAG_WRITE_FADVISE_DONTNEED = 1<<7, // write mode with LIBRADOS_OP_FLAG_FADVISE_DONTNEED
FLAG_NOSCRUB = 1<<8, // block periodic scrub
FLAG_NODEEP_SCRUB = 1<<9, // block periodic deep-scrub
FLAG_FULL_QUOTA = 1<<10, // pool is currently running out of quota, will set FLAG_FULL too
FLAG_NEARFULL = 1<<11, // pool is nearfull
FLAG_BACKFILLFULL = 1<<12, // pool is backfillfull
FLAG_SELFMANAGED_SNAPS = 1<<13, // pool uses selfmanaged snaps
FLAG_POOL_SNAPS = 1<<14, // pool has pool snaps
FLAG_CREATING = 1<<15, // initial pool PGs are being created
};
static const char *get_flag_name(int f) {
switch (f) {
case FLAG_HASHPSPOOL: return "hashpspool";
case FLAG_FULL: return "full";
case FLAG_EC_OVERWRITES: return "ec_overwrites";
case FLAG_INCOMPLETE_CLONES: return "incomplete_clones";
case FLAG_NODELETE: return "nodelete";
case FLAG_NOPGCHANGE: return "nopgchange";
case FLAG_NOSIZECHANGE: return "nosizechange";
case FLAG_WRITE_FADVISE_DONTNEED: return "write_fadvise_dontneed";
case FLAG_NOSCRUB: return "noscrub";
case FLAG_NODEEP_SCRUB: return "nodeep-scrub";
case FLAG_FULL_QUOTA: return "full_quota";
case FLAG_NEARFULL: return "nearfull";
case FLAG_BACKFILLFULL: return "backfillfull";
case FLAG_SELFMANAGED_SNAPS: return "selfmanaged_snaps";
case FLAG_POOL_SNAPS: return "pool_snaps";
case FLAG_CREATING: return "creating";
default: return "???";
}
}
static string get_flags_string(uint64_t f) {
string s;
for (unsigned n=0; f && n<64; ++n) {
if (f & (1ull << n)) {
if (s.length())
s += ",";
s += get_flag_name(1ull << n);
}
}
return s;
}
string get_flags_string() const {
return get_flags_string(flags);
}
static uint64_t get_flag_by_name(const string& name) {
if (name == "hashpspool")
return FLAG_HASHPSPOOL;
if (name == "full")
return FLAG_FULL;
if (name == "ec_overwrites")
return FLAG_EC_OVERWRITES;
if (name == "incomplete_clones")
return FLAG_INCOMPLETE_CLONES;
if (name == "nodelete")
return FLAG_NODELETE;
if (name == "nopgchange")
return FLAG_NOPGCHANGE;
if (name == "nosizechange")
return FLAG_NOSIZECHANGE;
if (name == "write_fadvise_dontneed")
return FLAG_WRITE_FADVISE_DONTNEED;
if (name == "noscrub")
return FLAG_NOSCRUB;
if (name == "nodeep-scrub")
return FLAG_NODEEP_SCRUB;
if (name == "full_quota")
return FLAG_FULL_QUOTA;
if (name == "nearfull")
return FLAG_NEARFULL;
if (name == "backfillfull")
return FLAG_BACKFILLFULL;
if (name == "selfmanaged_snaps")
return FLAG_SELFMANAGED_SNAPS;
if (name == "pool_snaps")
return FLAG_POOL_SNAPS;
if (name == "creating")
return FLAG_CREATING;
return 0;
}
/// converts the acting/up vector to a set of pg shards
void convert_to_pg_shards(const vector<int> &from, set<pg_shard_t>* to) const;
typedef enum {
CACHEMODE_NONE = 0, ///< no caching
CACHEMODE_WRITEBACK = 1, ///< write to cache, flush later
CACHEMODE_FORWARD = 2, ///< forward if not in cache
CACHEMODE_READONLY = 3, ///< handle reads, forward writes [not strongly consistent]
CACHEMODE_READFORWARD = 4, ///< forward reads, write to cache flush later
CACHEMODE_READPROXY = 5, ///< proxy reads, write to cache flush later
CACHEMODE_PROXY = 6, ///< proxy if not in cache
} cache_mode_t;
static const char *get_cache_mode_name(cache_mode_t m) {
switch (m) {
case CACHEMODE_NONE: return "none";
case CACHEMODE_WRITEBACK: return "writeback";
case CACHEMODE_FORWARD: return "forward";
case CACHEMODE_READONLY: return "readonly";
case CACHEMODE_READFORWARD: return "readforward";
case CACHEMODE_READPROXY: return "readproxy";
case CACHEMODE_PROXY: return "proxy";
default: return "unknown";
}
}
static cache_mode_t get_cache_mode_from_str(const string& s) {
if (s == "none")
return CACHEMODE_NONE;
if (s == "writeback")
return CACHEMODE_WRITEBACK;
if (s == "forward")
return CACHEMODE_FORWARD;
if (s == "readonly")
return CACHEMODE_READONLY;
if (s == "readforward")
return CACHEMODE_READFORWARD;
if (s == "readproxy")
return CACHEMODE_READPROXY;
if (s == "proxy")
return CACHEMODE_PROXY;
return (cache_mode_t)-1;
}
const char *get_cache_mode_name() const {
return get_cache_mode_name(cache_mode);
}
bool cache_mode_requires_hit_set() const {
switch (cache_mode) {
case CACHEMODE_NONE:
case CACHEMODE_FORWARD:
case CACHEMODE_READONLY:
case CACHEMODE_PROXY:
return false;
case CACHEMODE_WRITEBACK:
case CACHEMODE_READFORWARD:
case CACHEMODE_READPROXY:
return true;
default:
ceph_abort_msg("implement me");
}
}
enum {
PG_AUTOSCALE_MODE_OFF = 0,
PG_AUTOSCALE_MODE_WARN = 1,
PG_AUTOSCALE_MODE_ON = 2,
};
static const char *get_pg_autoscale_mode_name(int m) {
switch (m) {
case PG_AUTOSCALE_MODE_OFF: return "off";
case PG_AUTOSCALE_MODE_ON: return "on";
case PG_AUTOSCALE_MODE_WARN: return "warn";
default: return "???";
}
}
static int get_pg_autoscale_mode_by_name(const string& m) {
if (m == "off") {
return PG_AUTOSCALE_MODE_OFF;
}
if (m == "warn") {
return PG_AUTOSCALE_MODE_WARN;
}
if (m == "on") {
return PG_AUTOSCALE_MODE_ON;
}
return -1;
}
utime_t create_time;
uint64_t flags; ///< FLAG_*
__u8 type; ///< TYPE_*
__u8 size, min_size; ///< number of osds in each pg
__u8 crush_rule; ///< crush placement rule
__u8 object_hash; ///< hash mapping object name to ps
__u8 pg_autoscale_mode; ///< PG_AUTOSCALE_MODE_
private:
__u32 pg_num = 0, pgp_num = 0; ///< number of pgs
__u32 pg_num_pending = 0; ///< pg_num we are about to merge down to
__u32 pg_num_target = 0; ///< pg_num we should converge toward
__u32 pgp_num_target = 0; ///< pgp_num we should converge toward
public:
map<string,string> properties; ///< OBSOLETE
string erasure_code_profile; ///< name of the erasure code profile in OSDMap
epoch_t last_change; ///< most recent epoch changed, exclusing snapshot changes
/// last epoch that forced clients to resend
epoch_t last_force_op_resend = 0;
/// last epoch that forced clients to resend (pre-nautilus clients only)
epoch_t last_force_op_resend_prenautilus = 0;
/// last epoch that forced clients to resend (pre-luminous clients only)
epoch_t last_force_op_resend_preluminous = 0;
/// metadata for the most recent PG merge
pg_merge_meta_t last_pg_merge_meta;
snapid_t snap_seq; ///< seq for per-pool snapshot
epoch_t snap_epoch; ///< osdmap epoch of last snap
uint64_t auid; ///< who owns the pg
uint64_t quota_max_bytes; ///< maximum number of bytes for this pool
uint64_t quota_max_objects; ///< maximum number of objects for this pool
/*
* Pool snaps (global to this pool). These define a SnapContext for
* the pool, unless the client manually specifies an alternate
* context.
*/
map<snapid_t, pool_snap_info_t> snaps;
/*
* Alternatively, if we are defining non-pool snaps (e.g. via the
* Ceph MDS), we must track @removed_snaps (since @snaps is not
* used). Snaps and removed_snaps are to be used exclusive of each
* other!
*/
interval_set<snapid_t> removed_snaps;
unsigned pg_num_mask, pgp_num_mask;
set<uint64_t> tiers; ///< pools that are tiers of us
int64_t tier_of; ///< pool for which we are a tier
// Note that write wins for read+write ops
int64_t read_tier; ///< pool/tier for objecter to direct reads to
int64_t write_tier; ///< pool/tier for objecter to direct writes to
cache_mode_t cache_mode; ///< cache pool mode
bool is_tier() const { return tier_of >= 0; }
bool has_tiers() const { return !tiers.empty(); }
void clear_tier() {
tier_of = -1;
clear_read_tier();
clear_write_tier();
clear_tier_tunables();
}
bool has_read_tier() const { return read_tier >= 0; }
void clear_read_tier() { read_tier = -1; }
bool has_write_tier() const { return write_tier >= 0; }
void clear_write_tier() { write_tier = -1; }
void clear_tier_tunables() {
if (cache_mode != CACHEMODE_NONE)
flags |= FLAG_INCOMPLETE_CLONES;
cache_mode = CACHEMODE_NONE;
target_max_bytes = 0;
target_max_objects = 0;
cache_target_dirty_ratio_micro = 0;
cache_target_dirty_high_ratio_micro = 0;
cache_target_full_ratio_micro = 0;
hit_set_params = HitSet::Params();
hit_set_period = 0;
hit_set_count = 0;
hit_set_grade_decay_rate = 0;
hit_set_search_last_n = 0;
grade_table.resize(0);
}
uint64_t target_max_bytes; ///< tiering: target max pool size
uint64_t target_max_objects; ///< tiering: target max pool size
uint32_t cache_target_dirty_ratio_micro; ///< cache: fraction of target to leave dirty
uint32_t cache_target_dirty_high_ratio_micro; ///< cache: fraction of target to flush with high speed
uint32_t cache_target_full_ratio_micro; ///< cache: fraction of target to fill before we evict in earnest
uint32_t cache_min_flush_age; ///< minimum age (seconds) before we can flush
uint32_t cache_min_evict_age; ///< minimum age (seconds) before we can evict
HitSet::Params hit_set_params; ///< The HitSet params to use on this pool
uint32_t hit_set_period; ///< periodicity of HitSet segments (seconds)
uint32_t hit_set_count; ///< number of periods to retain
bool use_gmt_hitset; ///< use gmt to name the hitset archive object
uint32_t min_read_recency_for_promote; ///< minimum number of HitSet to check before promote on read
uint32_t min_write_recency_for_promote; ///< minimum number of HitSet to check before promote on write
uint32_t hit_set_grade_decay_rate; ///< current hit_set has highest priority on objects
///< temperature count,the follow hit_set's priority decay
///< by this params than pre hit_set
uint32_t hit_set_search_last_n; ///< accumulate atmost N hit_sets for temperature
uint32_t stripe_width; ///< erasure coded stripe size in bytes
uint64_t expected_num_objects; ///< expected number of objects on this pool, a value of 0 indicates
///< user does not specify any expected value
bool fast_read; ///< whether turn on fast read on the pool or not
pool_opts_t opts; ///< options
typedef enum {
TYPE_FINGERPRINT_NONE = 0,
TYPE_FINGERPRINT_SHA1 = 1,
} fingerprint_t;
static fingerprint_t get_fingerprint_from_str(const string& s) {
if (s == "none")
return TYPE_FINGERPRINT_NONE;
if (s == "sha1")
return TYPE_FINGERPRINT_SHA1;
return (fingerprint_t)-1;
}
const fingerprint_t get_fingerprint_type() const {
string fp_str;
opts.get(pool_opts_t::FINGERPRINT_ALGORITHM, &fp_str);
return get_fingerprint_from_str(fp_str);
}
const char *get_fingerprint_name() const {
string fp_str;
fingerprint_t fp_t;
opts.get(pool_opts_t::FINGERPRINT_ALGORITHM, &fp_str);
fp_t = get_fingerprint_from_str(fp_str);
return get_fingerprint_name(fp_t);
}
static const char *get_fingerprint_name(fingerprint_t m) {
switch (m) {
case TYPE_FINGERPRINT_NONE: return "none";
case TYPE_FINGERPRINT_SHA1: return "sha1";
default: return "unknown";
}
}
/// application -> key/value metadata
map<string, std::map<string, string>> application_metadata;
private:
vector<uint32_t> grade_table;
public:
uint32_t get_grade(unsigned i) const {
if (grade_table.size() <= i)
return 0;
return grade_table[i];
}
void calc_grade_table() {
unsigned v = 1000000;
grade_table.resize(hit_set_count);
for (unsigned i = 0; i < hit_set_count; i++) {
v = v * (1 - (hit_set_grade_decay_rate / 100.0));
grade_table[i] = v;
}
}
pg_pool_t()
: flags(0), type(0), size(0), min_size(0),
crush_rule(0), object_hash(0),
last_change(0),
snap_seq(0), snap_epoch(0),
auid(0),
quota_max_bytes(0), quota_max_objects(0),
pg_num_mask(0), pgp_num_mask(0),
tier_of(-1), read_tier(-1), write_tier(-1),
cache_mode(CACHEMODE_NONE),
target_max_bytes(0), target_max_objects(0),
cache_target_dirty_ratio_micro(0),
cache_target_dirty_high_ratio_micro(0),
cache_target_full_ratio_micro(0),
cache_min_flush_age(0),
cache_min_evict_age(0),
hit_set_params(),
hit_set_period(0),
hit_set_count(0),
use_gmt_hitset(true),
min_read_recency_for_promote(0),
min_write_recency_for_promote(0),
hit_set_grade_decay_rate(0),
hit_set_search_last_n(0),
stripe_width(0),
expected_num_objects(0),
fast_read(false),
opts()
{ }
void dump(Formatter *f) const;
const utime_t &get_create_time() const { return create_time; }
uint64_t get_flags() const { return flags; }
bool has_flag(uint64_t f) const { return flags & f; }
void set_flag(uint64_t f) { flags |= f; }
void unset_flag(uint64_t f) { flags &= ~f; }
bool require_rollback() const {
return is_erasure();
}
/// true if incomplete clones may be present
bool allow_incomplete_clones() const {
return cache_mode != CACHEMODE_NONE || has_flag(FLAG_INCOMPLETE_CLONES);
}
unsigned get_type() const { return type; }
unsigned get_size() const { return size; }
unsigned get_min_size() const { return min_size; }
int get_crush_rule() const { return crush_rule; }
int get_object_hash() const { return object_hash; }
const char *get_object_hash_name() const {
return ceph_str_hash_name(get_object_hash());
}
epoch_t get_last_change() const { return last_change; }
epoch_t get_last_force_op_resend() const { return last_force_op_resend; }
epoch_t get_last_force_op_resend_prenautilus() const {
return last_force_op_resend_prenautilus;
}
epoch_t get_last_force_op_resend_preluminous() const {
return last_force_op_resend_preluminous;
}
epoch_t get_snap_epoch() const { return snap_epoch; }
snapid_t get_snap_seq() const { return snap_seq; }
uint64_t get_auid() const { return auid; }
void set_snap_seq(snapid_t s) { snap_seq = s; }
void set_snap_epoch(epoch_t e) { snap_epoch = e; }
void set_stripe_width(uint32_t s) { stripe_width = s; }
uint32_t get_stripe_width() const { return stripe_width; }
bool is_replicated() const { return get_type() == TYPE_REPLICATED; }
bool is_erasure() const { return get_type() == TYPE_ERASURE; }
bool supports_omap() const {
return !(get_type() == TYPE_ERASURE);
}
bool requires_aligned_append() const {
return is_erasure() && !has_flag(FLAG_EC_OVERWRITES);
}
uint64_t required_alignment() const { return stripe_width; }
bool allows_ecoverwrites() const {
return has_flag(FLAG_EC_OVERWRITES);
}
bool can_shift_osds() const {
switch (get_type()) {
case TYPE_REPLICATED:
return true;
case TYPE_ERASURE:
return false;
default:
ceph_abort_msg("unhandled pool type");
}
}
unsigned get_pg_num() const { return pg_num; }
unsigned get_pgp_num() const { return pgp_num; }
unsigned get_pg_num_target() const { return pg_num_target; }
unsigned get_pgp_num_target() const { return pgp_num_target; }
unsigned get_pg_num_pending() const { return pg_num_pending; }
unsigned get_pg_num_mask() const { return pg_num_mask; }
unsigned get_pgp_num_mask() const { return pgp_num_mask; }
// if pg_num is not a multiple of two, pgs are not equally sized.
// return, for a given pg, the fraction (denominator) of the total
// pool size that it represents.
unsigned get_pg_num_divisor(pg_t pgid) const;
bool is_pending_merge(pg_t pgid, bool *target) const;
void set_pg_num(int p) {
pg_num = p;
pg_num_pending = p;
calc_pg_masks();
}
void set_pgp_num(int p) {
pgp_num = p;
calc_pg_masks();
}
void set_pg_num_pending(int p) {
pg_num_pending = p;
calc_pg_masks();
}
void set_pg_num_target(int p) {
pg_num_target = p;
}
void set_pgp_num_target(int p) {
pgp_num_target = p;
}
void dec_pg_num(pg_t source_pgid,
epoch_t ready_epoch,
eversion_t source_version,
eversion_t target_version,
epoch_t last_epoch_started,
epoch_t last_epoch_clean) {
--pg_num;
last_pg_merge_meta.source_pgid = source_pgid;
last_pg_merge_meta.ready_epoch = ready_epoch;
last_pg_merge_meta.source_version = source_version;
last_pg_merge_meta.target_version = target_version;
last_pg_merge_meta.last_epoch_started = last_epoch_started;
last_pg_merge_meta.last_epoch_clean = last_epoch_clean;
calc_pg_masks();
}
void set_quota_max_bytes(uint64_t m) {
quota_max_bytes = m;
}
uint64_t get_quota_max_bytes() {
return quota_max_bytes;
}
void set_quota_max_objects(uint64_t m) {
quota_max_objects = m;
}
uint64_t get_quota_max_objects() {
return quota_max_objects;
}
void set_last_force_op_resend(uint64_t t) {
last_force_op_resend = t;
last_force_op_resend_prenautilus = t;
last_force_op_resend_preluminous = t;
}
void calc_pg_masks();
/*
* we have two snap modes:
* - pool global snaps
* - snap existence/non-existence defined by snaps[] and snap_seq
* - user managed snaps
* - removal governed by removed_snaps
*
* we know which mode we're using based on whether removed_snaps is empty.
* If nothing has been created, both functions report false.
*/
bool is_pool_snaps_mode() const;
bool is_unmanaged_snaps_mode() const;
bool is_removed_snap(snapid_t s) const;
/*
* build set of known-removed sets from either pool snaps or
* explicit removed_snaps set.
*/
void build_removed_snaps(interval_set<snapid_t>& rs) const;
bool maybe_updated_removed_snaps(const interval_set<snapid_t>& cached) const;
snapid_t snap_exists(const char *s) const;
void add_snap(const char *n, utime_t stamp);
void add_unmanaged_snap(uint64_t& snapid);
void remove_snap(snapid_t s);
void remove_unmanaged_snap(snapid_t s);
SnapContext get_snap_context() const;
/// hash a object name+namespace key to a hash position
uint32_t hash_key(const string& key, const string& ns) const;
/// round a hash position down to a pg num
uint32_t raw_hash_to_pg(uint32_t v) const;
/*
* map a raw pg (with full precision ps) into an actual pg, for storage
*/
pg_t raw_pg_to_pg(pg_t pg) const;
/*
* map raw pg (full precision ps) into a placement seed. include
* pool id in that value so that different pools don't use the same
* seeds.
*/
ps_t raw_pg_to_pps(pg_t pg) const;
/// choose a random hash position within a pg
uint32_t get_random_pg_position(pg_t pgid, uint32_t seed) const;
void encode(bufferlist& bl, uint64_t features) const;
void decode(bufferlist::const_iterator& bl);
static void generate_test_instances(list<pg_pool_t*>& o);
};
WRITE_CLASS_ENCODER_FEATURES(pg_pool_t)
ostream& operator<<(ostream& out, const pg_pool_t& p);
/**
* a summation of object stats
*
* This is just a container for object stats; we don't know what for.
*
* If you add members in object_stat_sum_t, you should make sure there are
* not padding among these members.
* You should also modify the padding_check function.
*/
struct object_stat_sum_t {
/**************************************************************************
* WARNING: be sure to update operator==, floor, and split when
* adding/removing fields!
**************************************************************************/
int64_t num_bytes; // in bytes
int64_t num_objects;
int64_t num_object_clones;
int64_t num_object_copies; // num_objects * num_replicas
int64_t num_objects_missing_on_primary;
int64_t num_objects_degraded;
int64_t num_objects_unfound;
int64_t num_rd;
int64_t num_rd_kb;
int64_t num_wr;
int64_t num_wr_kb;
int64_t num_scrub_errors; // total deep and shallow scrub errors
int64_t num_objects_recovered;
int64_t num_bytes_recovered;
int64_t num_keys_recovered;
int64_t num_shallow_scrub_errors;
int64_t num_deep_scrub_errors;
int64_t num_objects_dirty;
int64_t num_whiteouts;
int64_t num_objects_omap;
int64_t num_objects_hit_set_archive;
int64_t num_objects_misplaced;
int64_t num_bytes_hit_set_archive;
int64_t num_flush;
int64_t num_flush_kb;
int64_t num_evict;
int64_t num_evict_kb;
int64_t num_promote;
int32_t num_flush_mode_high; // 1 when in high flush mode, otherwise 0
int32_t num_flush_mode_low; // 1 when in low flush mode, otherwise 0
int32_t num_evict_mode_some; // 1 when in evict some mode, otherwise 0
int32_t num_evict_mode_full; // 1 when in evict full mode, otherwise 0
int64_t num_objects_pinned;
int64_t num_objects_missing;
int64_t num_legacy_snapsets; ///< upper bound on pre-luminous-style SnapSets
int64_t num_large_omap_objects = 0;
int64_t num_objects_manifest = 0;
int64_t num_omap_bytes = 0;
int64_t num_omap_keys = 0;
int64_t num_objects_repaired = 0;
object_stat_sum_t()
: num_bytes(0),
num_objects(0), num_object_clones(0), num_object_copies(0),
num_objects_missing_on_primary(0), num_objects_degraded(0),
num_objects_unfound(0),
num_rd(0), num_rd_kb(0), num_wr(0), num_wr_kb(0),
num_scrub_errors(0),
num_objects_recovered(0),
num_bytes_recovered(0),
num_keys_recovered(0),
num_shallow_scrub_errors(0),
num_deep_scrub_errors(0),
num_objects_dirty(0),
num_whiteouts(0),
num_objects_omap(0),
num_objects_hit_set_archive(0),
num_objects_misplaced(0),
num_bytes_hit_set_archive(0),
num_flush(0),
num_flush_kb(0),
num_evict(0),
num_evict_kb(0),
num_promote(0),
num_flush_mode_high(0), num_flush_mode_low(0),
num_evict_mode_some(0), num_evict_mode_full(0),
num_objects_pinned(0),
num_objects_missing(0),
num_legacy_snapsets(0)
{}
void floor(int64_t f) {
#define FLOOR(x) if (x < f) x = f
FLOOR(num_bytes);
FLOOR(num_objects);
FLOOR(num_object_clones);
FLOOR(num_object_copies);
FLOOR(num_objects_missing_on_primary);
FLOOR(num_objects_missing);
FLOOR(num_objects_degraded);
FLOOR(num_objects_misplaced);
FLOOR(num_objects_unfound);
FLOOR(num_rd);
FLOOR(num_rd_kb);
FLOOR(num_wr);
FLOOR(num_wr_kb);
FLOOR(num_large_omap_objects);
FLOOR(num_objects_manifest);
FLOOR(num_omap_bytes);
FLOOR(num_omap_keys);
FLOOR(num_shallow_scrub_errors);
FLOOR(num_deep_scrub_errors);
num_scrub_errors = num_shallow_scrub_errors + num_deep_scrub_errors;
FLOOR(num_objects_recovered);
FLOOR(num_bytes_recovered);
FLOOR(num_keys_recovered);
FLOOR(num_objects_dirty);
FLOOR(num_whiteouts);
FLOOR(num_objects_omap);
FLOOR(num_objects_hit_set_archive);
FLOOR(num_bytes_hit_set_archive);
FLOOR(num_flush);
FLOOR(num_flush_kb);
FLOOR(num_evict);
FLOOR(num_evict_kb);
FLOOR(num_promote);
FLOOR(num_flush_mode_high);
FLOOR(num_flush_mode_low);
FLOOR(num_evict_mode_some);
FLOOR(num_evict_mode_full);
FLOOR(num_objects_pinned);
FLOOR(num_legacy_snapsets);
FLOOR(num_objects_repaired);
#undef FLOOR
}
void split(vector<object_stat_sum_t> &out) const {
#define SPLIT(PARAM) \
for (unsigned i = 0; i < out.size(); ++i) { \
out[i].PARAM = PARAM / out.size(); \
if (i < (PARAM % out.size())) { \
out[i].PARAM++; \
} \
}
#define SPLIT_PRESERVE_NONZERO(PARAM) \
for (unsigned i = 0; i < out.size(); ++i) { \
if (PARAM) \
out[i].PARAM = 1 + PARAM / out.size(); \
else \
out[i].PARAM = 0; \
}
SPLIT(num_bytes);
SPLIT(num_objects);
SPLIT(num_object_clones);
SPLIT(num_object_copies);
SPLIT(num_objects_missing_on_primary);
SPLIT(num_objects_missing);
SPLIT(num_objects_degraded);
SPLIT(num_objects_misplaced);
SPLIT(num_objects_unfound);
SPLIT(num_rd);
SPLIT(num_rd_kb);
SPLIT(num_wr);
SPLIT(num_wr_kb);
SPLIT(num_large_omap_objects);
SPLIT(num_objects_manifest);
SPLIT(num_omap_bytes);
SPLIT(num_omap_keys);
SPLIT(num_objects_repaired);
SPLIT_PRESERVE_NONZERO(num_shallow_scrub_errors);
SPLIT_PRESERVE_NONZERO(num_deep_scrub_errors);
for (unsigned i = 0; i < out.size(); ++i) {
out[i].num_scrub_errors = out[i].num_shallow_scrub_errors +
out[i].num_deep_scrub_errors;
}
SPLIT(num_objects_recovered);
SPLIT(num_bytes_recovered);
SPLIT(num_keys_recovered);
SPLIT(num_objects_dirty);
SPLIT(num_whiteouts);
SPLIT(num_objects_omap);
SPLIT(num_objects_hit_set_archive);
SPLIT(num_bytes_hit_set_archive);
SPLIT(num_flush);
SPLIT(num_flush_kb);
SPLIT(num_evict);
SPLIT(num_evict_kb);
SPLIT(num_promote);
SPLIT(num_flush_mode_high);
SPLIT(num_flush_mode_low);
SPLIT(num_evict_mode_some);
SPLIT(num_evict_mode_full);
SPLIT(num_objects_pinned);
SPLIT_PRESERVE_NONZERO(num_legacy_snapsets);
#undef SPLIT
#undef SPLIT_PRESERVE_NONZERO
}
void clear() {
// FIPS zeroization audit 20191117: this memset is not security related.
memset(this, 0, sizeof(*this));
}
void calc_copies(int nrep) {
num_object_copies = nrep * num_objects;
}
bool is_zero() const {
return mem_is_zero((char*)this, sizeof(*this));
}
void add(const object_stat_sum_t& o);
void sub(const object_stat_sum_t& o);
void dump(Formatter *f) const;
void padding_check() {
static_assert(
sizeof(object_stat_sum_t) ==
sizeof(num_bytes) +
sizeof(num_objects) +
sizeof(num_object_clones) +
sizeof(num_object_copies) +
sizeof(num_objects_missing_on_primary) +
sizeof(num_objects_degraded) +
sizeof(num_objects_unfound) +
sizeof(num_rd) +
sizeof(num_rd_kb) +
sizeof(num_wr) +
sizeof(num_wr_kb) +
sizeof(num_scrub_errors) +
sizeof(num_large_omap_objects) +
sizeof(num_objects_manifest) +
sizeof(num_omap_bytes) +
sizeof(num_omap_keys) +
sizeof(num_objects_repaired) +
sizeof(num_objects_recovered) +
sizeof(num_bytes_recovered) +
sizeof(num_keys_recovered) +
sizeof(num_shallow_scrub_errors) +
sizeof(num_deep_scrub_errors) +
sizeof(num_objects_dirty) +
sizeof(num_whiteouts) +
sizeof(num_objects_omap) +
sizeof(num_objects_hit_set_archive) +
sizeof(num_objects_misplaced) +
sizeof(num_bytes_hit_set_archive) +
sizeof(num_flush) +
sizeof(num_flush_kb) +
sizeof(num_evict) +
sizeof(num_evict_kb) +
sizeof(num_promote) +
sizeof(num_flush_mode_high) +
sizeof(num_flush_mode_low) +
sizeof(num_evict_mode_some) +
sizeof(num_evict_mode_full) +
sizeof(num_objects_pinned) +
sizeof(num_objects_missing) +
sizeof(num_legacy_snapsets)
,
"object_stat_sum_t have padding");
}
void encode(bufferlist& bl) const;
void decode(bufferlist::const_iterator& bl);
static void generate_test_instances(list<object_stat_sum_t*>& o);
};
WRITE_CLASS_ENCODER(object_stat_sum_t)
bool operator==(const object_stat_sum_t& l, const object_stat_sum_t& r);
/**
* a collection of object stat sums
*
* This is a collection of stat sums over different categories.
*/
struct object_stat_collection_t {
/**************************************************************************
* WARNING: be sure to update the operator== when adding/removing fields! *
**************************************************************************/
object_stat_sum_t sum;
void calc_copies(int nrep) {
sum.calc_copies(nrep);
}
void dump(Formatter *f) const;
void encode(bufferlist& bl) const;
void decode(bufferlist::const_iterator& bl);
static void generate_test_instances(list<object_stat_collection_t*>& o);
bool is_zero() const {
return sum.is_zero();
}
void clear() {
sum.clear();
}
void floor(int64_t f) {
sum.floor(f);
}
void add(const object_stat_sum_t& o) {
sum.add(o);
}
void add(const object_stat_collection_t& o) {
sum.add(o.sum);
}
void sub(const object_stat_collection_t& o) {
sum.sub(o.sum);
}
};
WRITE_CLASS_ENCODER(object_stat_collection_t)
inline bool operator==(const object_stat_collection_t& l,
const object_stat_collection_t& r) {
return l.sum == r.sum;
}
/** pg_stat
* aggregate stats for a single PG.
*/
struct pg_stat_t {
/**************************************************************************
* WARNING: be sure to update the operator== when adding/removing fields! *
**************************************************************************/
eversion_t version;
version_t reported_seq; // sequence number
epoch_t reported_epoch; // epoch of this report
uint64_t state;
utime_t last_fresh; // last reported
utime_t last_change; // new state != previous state
utime_t last_active; // state & PG_STATE_ACTIVE
utime_t last_peered; // state & PG_STATE_ACTIVE || state & PG_STATE_PEERED
utime_t last_clean; // state & PG_STATE_CLEAN
utime_t last_unstale; // (state & PG_STATE_STALE) == 0
utime_t last_undegraded; // (state & PG_STATE_DEGRADED) == 0
utime_t last_fullsized; // (state & PG_STATE_UNDERSIZED) == 0
eversion_t log_start; // (log_start,version]
eversion_t ondisk_log_start; // there may be more on disk
epoch_t created;
epoch_t last_epoch_clean;
pg_t parent;
__u32 parent_split_bits;
eversion_t last_scrub;
eversion_t last_deep_scrub;
utime_t last_scrub_stamp;
utime_t last_deep_scrub_stamp;
utime_t last_clean_scrub_stamp;
object_stat_collection_t stats;
int64_t log_size;
int64_t ondisk_log_size; // >= active_log_size
vector<int32_t> up, acting;
vector<pg_shard_t> avail_no_missing;
map< std::set<pg_shard_t>, int32_t > object_location_counts;
epoch_t mapping_epoch;
vector<int32_t> blocked_by; ///< osds on which the pg is blocked
interval_set<snapid_t> purged_snaps; ///< recently removed snaps that we've purged
utime_t last_became_active;
utime_t last_became_peered;
/// up, acting primaries
int32_t up_primary;
int32_t acting_primary;
// snaptrimq.size() is 64bit, but let's be serious - anything over 50k is
// absurd already, so cap it to 2^32 and save 4 bytes at the same time
uint32_t snaptrimq_len;
bool stats_invalid:1;
/// true if num_objects_dirty is not accurate (because it was not
/// maintained starting from pool creation)
bool dirty_stats_invalid:1;
bool omap_stats_invalid:1;
bool hitset_stats_invalid:1;
bool hitset_bytes_stats_invalid:1;
bool pin_stats_invalid:1;
bool manifest_stats_invalid:1;
pg_stat_t()
: reported_seq(0),
reported_epoch(0),
state(0),
created(0), last_epoch_clean(0),
parent_split_bits(0),
log_size(0), ondisk_log_size(0),
mapping_epoch(0),
up_primary(-1),
acting_primary(-1),
snaptrimq_len(0),
stats_invalid(false),
dirty_stats_invalid(false),
omap_stats_invalid(false),
hitset_stats_invalid(false),
hitset_bytes_stats_invalid(false),
pin_stats_invalid(false),
manifest_stats_invalid(false)
{ }
epoch_t get_effective_last_epoch_clean() const {
if (state & PG_STATE_CLEAN) {
// we are clean as of this report, and should thus take the
// reported epoch
return reported_epoch;
} else {
return last_epoch_clean;
}
}
pair<epoch_t, version_t> get_version_pair() const {
return make_pair(reported_epoch, reported_seq);
}
void floor(int64_t f) {
stats.floor(f);
if (log_size < f)
log_size = f;
if (ondisk_log_size < f)
ondisk_log_size = f;
if (snaptrimq_len < f)
snaptrimq_len = f;
}
void add_sub_invalid_flags(const pg_stat_t& o) {
// adding (or subtracting!) invalid stats render our stats invalid too
stats_invalid |= o.stats_invalid;
dirty_stats_invalid |= o.dirty_stats_invalid;
omap_stats_invalid |= o.omap_stats_invalid;
hitset_stats_invalid |= o.hitset_stats_invalid;
hitset_bytes_stats_invalid |= o.hitset_bytes_stats_invalid;
pin_stats_invalid |= o.pin_stats_invalid;
manifest_stats_invalid |= o.manifest_stats_invalid;
}
void add(const pg_stat_t& o) {
stats.add(o.stats);
log_size += o.log_size;
ondisk_log_size += o.ondisk_log_size;
snaptrimq_len = std::min((uint64_t)snaptrimq_len + o.snaptrimq_len,
(uint64_t)(1ull << 31));
add_sub_invalid_flags(o);
}
void sub(const pg_stat_t& o) {
stats.sub(o.stats);
log_size -= o.log_size;
ondisk_log_size -= o.ondisk_log_size;
if (o.snaptrimq_len < snaptrimq_len) {
snaptrimq_len -= o.snaptrimq_len;
} else {
snaptrimq_len = 0;
}
add_sub_invalid_flags(o);
}
bool is_acting_osd(int32_t osd, bool primary) const;
void dump(Formatter *f) const;
void dump_brief(Formatter *f) const;
void encode(bufferlist &bl) const;
void decode(bufferlist::const_iterator &bl);
static void generate_test_instances(list<pg_stat_t*>& o);
};
WRITE_CLASS_ENCODER(pg_stat_t)
bool operator==(const pg_stat_t& l, const pg_stat_t& r);
/** store_statfs_t
* ObjectStore full statfs information
*/
struct store_statfs_t
{
uint64_t total = 0; ///< Total bytes
uint64_t available = 0; ///< Free bytes available
uint64_t internally_reserved = 0; ///< Bytes reserved for internal purposes
int64_t allocated = 0; ///< Bytes allocated by the store
int64_t data_stored = 0; ///< Bytes actually stored by the user
int64_t data_compressed = 0; ///< Bytes stored after compression
int64_t data_compressed_allocated = 0; ///< Bytes allocated for compressed data
int64_t data_compressed_original = 0; ///< Bytes that were compressed
int64_t omap_allocated = 0; ///< approx usage of omap data
int64_t internal_metadata = 0; ///< approx usage of internal metadata
void reset() {
*this = store_statfs_t();
}
void floor(int64_t f) {
#define FLOOR(x) if (int64_t(x) < f) x = f
FLOOR(total);
FLOOR(available);
FLOOR(internally_reserved);
FLOOR(allocated);
FLOOR(data_stored);
FLOOR(data_compressed);
FLOOR(data_compressed_allocated);
FLOOR(data_compressed_original);
FLOOR(omap_allocated);
FLOOR(internal_metadata);
#undef FLOOR
}
bool operator ==(const store_statfs_t& other) const;
bool is_zero() const {
return *this == store_statfs_t();
}
uint64_t get_used() const {
return total - available - internally_reserved;
}
// this accumulates both actually used and statfs's internally_reserved
uint64_t get_used_raw() const {
return total - available;
}
float get_used_raw_ratio() const {
if (total) {
return (float)get_used_raw() / (float)total;
} else {
return 0.0;
}
}
// helpers to ease legacy code porting
uint64_t kb_avail() const {
return available >> 10;
}
uint64_t kb() const {
return total >> 10;
}
uint64_t kb_used() const {
return (total - available - internally_reserved) >> 10;
}
uint64_t kb_used_raw() const {
return get_used_raw() >> 10;
}
uint64_t kb_used_data() const {
return allocated >> 10;
}
uint64_t kb_used_omap() const {
return omap_allocated >> 10;
}
uint64_t kb_used_internal_metadata() const {
return internal_metadata >> 10;
}
void add(const store_statfs_t& o) {
total += o.total;
available += o.available;
internally_reserved += o.internally_reserved;
allocated += o.allocated;
data_stored += o.data_stored;
data_compressed += o.data_compressed;
data_compressed_allocated += o.data_compressed_allocated;
data_compressed_original += o.data_compressed_original;
omap_allocated += o.omap_allocated;
internal_metadata += o.internal_metadata;
}
void sub(const store_statfs_t& o) {
total -= o.total;
available -= o.available;
internally_reserved -= o.internally_reserved;
allocated -= o.allocated;
data_stored -= o.data_stored;
data_compressed -= o.data_compressed;
data_compressed_allocated -= o.data_compressed_allocated;
data_compressed_original -= o.data_compressed_original;
omap_allocated -= o.omap_allocated;
internal_metadata -= o.internal_metadata;
}
void dump(Formatter *f) const;
DENC(store_statfs_t, v, p) {
DENC_START(1, 1, p);
denc(v.total, p);
denc(v.available, p);
denc(v.internally_reserved, p);
denc(v.allocated, p);
denc(v.data_stored, p);
denc(v.data_compressed, p);
denc(v.data_compressed_allocated, p);
denc(v.data_compressed_original, p);
denc(v.omap_allocated, p);
denc(v.internal_metadata, p);
DENC_FINISH(p);
}
static void generate_test_instances(list<store_statfs_t*>& o);
};
WRITE_CLASS_DENC(store_statfs_t)
ostream &operator<<(ostream &lhs, const store_statfs_t &rhs);
/** osd_stat
* aggregate stats for an osd
*/
struct osd_stat_t {
store_statfs_t statfs;
vector<int> hb_peers;
int32_t snap_trim_queue_len, num_snap_trimming;
uint64_t num_shards_repaired;
pow2_hist_t op_queue_age_hist;
objectstore_perf_stat_t os_perf_stat;
osd_alerts_t os_alerts;
epoch_t up_from = 0;
uint64_t seq = 0;
uint32_t num_pgs = 0;
uint32_t num_osds = 0;
uint32_t num_per_pool_osds = 0;
struct Interfaces {
uint32_t last_update; // in seconds
uint32_t back_pingtime[3];
uint32_t back_min[3];
uint32_t back_max[3];
uint32_t back_last;
uint32_t front_pingtime[3];
uint32_t front_min[3];
uint32_t front_max[3];
uint32_t front_last;
};
map<int, Interfaces> hb_pingtime; ///< map of osd id to Interfaces
osd_stat_t() : snap_trim_queue_len(0), num_snap_trimming(0),
num_shards_repaired(0) {}
void add(const osd_stat_t& o) {
statfs.add(o.statfs);
snap_trim_queue_len += o.snap_trim_queue_len;
num_snap_trimming += o.num_snap_trimming;
num_shards_repaired += o.num_shards_repaired;
op_queue_age_hist.add(o.op_queue_age_hist);
os_perf_stat.add(o.os_perf_stat);
num_pgs += o.num_pgs;
num_osds += o.num_osds;
num_per_pool_osds += o.num_per_pool_osds;
for (const auto& a : o.os_alerts) {
auto& target = os_alerts[a.first];
for (auto& i : a.second) {
target.emplace(i.first, i.second);
}
}
}
void sub(const osd_stat_t& o) {
statfs.sub(o.statfs);
snap_trim_queue_len -= o.snap_trim_queue_len;
num_snap_trimming -= o.num_snap_trimming;
num_shards_repaired -= o.num_shards_repaired;
op_queue_age_hist.sub(o.op_queue_age_hist);
os_perf_stat.sub(o.os_perf_stat);
num_pgs -= o.num_pgs;
num_osds -= o.num_osds;
num_per_pool_osds -= o.num_per_pool_osds;
for (const auto& a : o.os_alerts) {
auto& target = os_alerts[a.first];
for (auto& i : a.second) {
target.erase(i.first);
}
if (target.empty()) {
os_alerts.erase(a.first);
}
}
}
void dump(Formatter *f, bool with_net = true) const;
void dump_ping_time(Formatter *f) const;
void encode(bufferlist &bl, uint64_t features) const;
void decode(bufferlist::const_iterator &bl);
static void generate_test_instances(std::list<osd_stat_t*>& o);
};
WRITE_CLASS_ENCODER_FEATURES(osd_stat_t)
inline bool operator==(const osd_stat_t& l, const osd_stat_t& r) {
return l.statfs == r.statfs &&
l.snap_trim_queue_len == r.snap_trim_queue_len &&
l.num_snap_trimming == r.num_snap_trimming &&
l.num_shards_repaired == r.num_shards_repaired &&
l.hb_peers == r.hb_peers &&
l.op_queue_age_hist == r.op_queue_age_hist &&
l.os_perf_stat == r.os_perf_stat &&
l.num_pgs == r.num_pgs &&
l.num_osds == r.num_osds &&
l.num_per_pool_osds == r.num_per_pool_osds;
}
inline bool operator!=(const osd_stat_t& l, const osd_stat_t& r) {
return !(l == r);
}
inline ostream& operator<<(ostream& out, const osd_stat_t& s) {
return out << "osd_stat(" << s.statfs << ", "
<< "peers " << s.hb_peers
<< " op hist " << s.op_queue_age_hist.h
<< ")";
}
/*
* summation over an entire pool
*/
struct pool_stat_t {
object_stat_collection_t stats;
store_statfs_t store_stats;
int64_t log_size;
int64_t ondisk_log_size; // >= active_log_size
int32_t up; ///< number of up replicas or shards
int32_t acting; ///< number of acting replicas or shards
int32_t num_store_stats; ///< amount of store_stats accumulated
pool_stat_t() : log_size(0), ondisk_log_size(0), up(0), acting(0),
num_store_stats(0)
{ }
void floor(int64_t f) {
stats.floor(f);
store_stats.floor(f);
if (log_size < f)
log_size = f;
if (ondisk_log_size < f)
ondisk_log_size = f;
if (up < f)
up = f;
if (acting < f)
acting = f;
if (num_store_stats < f)
num_store_stats = f;
}
void add(const store_statfs_t& o) {
store_stats.add(o);
++num_store_stats;
}
void sub(const store_statfs_t& o) {
store_stats.sub(o);
--num_store_stats;
}
void add(const pg_stat_t& o) {
stats.add(o.stats);
log_size += o.log_size;
ondisk_log_size += o.ondisk_log_size;
up += o.up.size();
acting += o.acting.size();
}
void sub(const pg_stat_t& o) {
stats.sub(o.stats);
log_size -= o.log_size;
ondisk_log_size -= o.ondisk_log_size;
up -= o.up.size();
acting -= o.acting.size();
}
bool is_zero() const {
return (stats.is_zero() &&
store_stats.is_zero() &&
log_size == 0 &&
ondisk_log_size == 0 &&
up == 0 &&
acting == 0 &&
num_store_stats == 0);
}
// helper accessors to retrieve used/netto bytes depending on the
// collection method: new per-pool objectstore report or legacy PG
// summation at OSD.
// In legacy mode used and netto values are the same. But for new per-pool
// collection 'used' provides amount of space ALLOCATED at all related OSDs
// and 'netto' is amount of stored user data.
uint64_t get_allocated_bytes(bool per_pool) const {
uint64_t allocated_bytes;
if (per_pool) {
allocated_bytes = store_stats.allocated;
} else {
// legacy mode, use numbers from 'stats'
allocated_bytes = stats.sum.num_bytes +
stats.sum.num_bytes_hit_set_archive;
}
// omap is not broken out by pool by nautilus bluestore
allocated_bytes += stats.sum.num_omap_bytes;
return allocated_bytes;
}
uint64_t get_user_bytes(float raw_used_rate, bool per_pool) const {
uint64_t user_bytes;
if (per_pool) {
user_bytes = raw_used_rate ? store_stats.data_stored / raw_used_rate : 0;
} else {
// legacy mode, use numbers from 'stats'
user_bytes = stats.sum.num_bytes +
stats.sum.num_bytes_hit_set_archive;
}
// omap is not broken out by pool by nautilus bluestore
user_bytes += stats.sum.num_omap_bytes;
return user_bytes;
}
void dump(Formatter *f) const;
void encode(bufferlist &bl, uint64_t features) const;
void decode(bufferlist::const_iterator &bl);
static void generate_test_instances(list<pool_stat_t*>& o);
};
WRITE_CLASS_ENCODER_FEATURES(pool_stat_t)
// -----------------------------------------
/**
* pg_hit_set_info_t - information about a single recorded HitSet
*
* Track basic metadata about a HitSet, like the number of insertions
* and the time range it covers.
*/
struct pg_hit_set_info_t {
utime_t begin, end; ///< time interval
eversion_t version; ///< version this HitSet object was written
bool using_gmt; ///< use gmt for creating the hit_set archive object name
friend bool operator==(const pg_hit_set_info_t& l,
const pg_hit_set_info_t& r) {
return
l.begin == r.begin &&
l.end == r.end &&
l.version == r.version &&
l.using_gmt == r.using_gmt;
}
explicit pg_hit_set_info_t(bool using_gmt = true)
: using_gmt(using_gmt) {}
void encode(bufferlist &bl) const;
void decode(bufferlist::const_iterator &bl);
void dump(Formatter *f) const;
static void generate_test_instances(list<pg_hit_set_info_t*>& o);
};
WRITE_CLASS_ENCODER(pg_hit_set_info_t)
/**
* pg_hit_set_history_t - information about a history of hitsets
*
* Include information about the currently accumulating hit set as well
* as archived/historical ones.
*/
struct pg_hit_set_history_t {
eversion_t current_last_update; ///< last version inserted into current set
list<pg_hit_set_info_t> history; ///< archived sets, sorted oldest -> newest
friend bool operator==(const pg_hit_set_history_t& l,
const pg_hit_set_history_t& r) {
return
l.current_last_update == r.current_last_update &&
l.history == r.history;
}
void encode(bufferlist &bl) const;
void decode(bufferlist::const_iterator &bl);
void dump(Formatter *f) const;
static void generate_test_instances(list<pg_hit_set_history_t*>& o);
};
WRITE_CLASS_ENCODER(pg_hit_set_history_t)
// -----------------------------------------
/**
* pg_history_t - information about recent pg peering/mapping history
*
* This is aggressively shared between OSDs to bound the amount of past
* history they need to worry about.
*/
struct pg_history_t {
epoch_t epoch_created; // epoch in which *pg* was created (pool or pg)
epoch_t epoch_pool_created; // epoch in which *pool* was created
// (note: may be pg creation epoch for
// pre-luminous clusters)
epoch_t last_epoch_started; // lower bound on last epoch started (anywhere, not necessarily locally)
epoch_t last_interval_started; // first epoch of last_epoch_started interval
epoch_t last_epoch_clean; // lower bound on last epoch the PG was completely clean.
epoch_t last_interval_clean; // first epoch of last_epoch_clean interval
epoch_t last_epoch_split; // as parent or child
epoch_t last_epoch_marked_full; // pool or cluster
/**
* In the event of a map discontinuity, same_*_since may reflect the first
* map the osd has seen in the new map sequence rather than the actual start
* of the interval. This is ok since a discontinuity at epoch e means there
* must have been a clean interval between e and now and that we cannot be
* in the active set during the interval containing e.
*/
epoch_t same_up_since; // same acting set since
epoch_t same_interval_since; // same acting AND up set since
epoch_t same_primary_since; // same primary at least back through this epoch.
eversion_t last_scrub;
eversion_t last_deep_scrub;
utime_t last_scrub_stamp;
utime_t last_deep_scrub_stamp;
utime_t last_clean_scrub_stamp;
friend bool operator==(const pg_history_t& l, const pg_history_t& r) {
return
l.epoch_created == r.epoch_created &&
l.epoch_pool_created == r.epoch_pool_created &&
l.last_epoch_started == r.last_epoch_started &&
l.last_interval_started == r.last_interval_started &&
l.last_epoch_clean == r.last_epoch_clean &&
l.last_interval_clean == r.last_interval_clean &&
l.last_epoch_split == r.last_epoch_split &&
l.last_epoch_marked_full == r.last_epoch_marked_full &&
l.same_up_since == r.same_up_since &&
l.same_interval_since == r.same_interval_since &&
l.same_primary_since == r.same_primary_since &&
l.last_scrub == r.last_scrub &&
l.last_deep_scrub == r.last_deep_scrub &&
l.last_scrub_stamp == r.last_scrub_stamp &&
l.last_deep_scrub_stamp == r.last_deep_scrub_stamp &&
l.last_clean_scrub_stamp == r.last_clean_scrub_stamp;
}
pg_history_t()
: epoch_created(0),
epoch_pool_created(0),
last_epoch_started(0),
last_interval_started(0),
last_epoch_clean(0),
last_interval_clean(0),
last_epoch_split(0),
last_epoch_marked_full(0),
same_up_since(0), same_interval_since(0), same_primary_since(0) {}
bool merge(const pg_history_t &other) {
// Here, we only update the fields which cannot be calculated from the OSDmap.
bool modified = false;
if (epoch_created < other.epoch_created) {
epoch_created = other.epoch_created;
modified = true;
}
if (epoch_pool_created < other.epoch_pool_created) {
// FIXME: for jewel compat only; this should either be 0 or always the
// same value across all pg instances.
epoch_pool_created = other.epoch_pool_created;
modified = true;
}
if (last_epoch_started < other.last_epoch_started) {
last_epoch_started = other.last_epoch_started;
modified = true;
}
if (last_interval_started < other.last_interval_started) {
last_interval_started = other.last_interval_started;
modified = true;
}
if (last_epoch_clean < other.last_epoch_clean) {
last_epoch_clean = other.last_epoch_clean;
modified = true;
}
if (last_interval_clean < other.last_interval_clean) {
last_interval_clean = other.last_interval_clean;
modified = true;
}
if (last_epoch_split < other.last_epoch_split) {
last_epoch_split = other.last_epoch_split;
modified = true;
}
if (last_epoch_marked_full < other.last_epoch_marked_full) {
last_epoch_marked_full = other.last_epoch_marked_full;
modified = true;
}
if (other.last_scrub > last_scrub) {
last_scrub = other.last_scrub;
modified = true;
}
if (other.last_scrub_stamp > last_scrub_stamp) {
last_scrub_stamp = other.last_scrub_stamp;
modified = true;
}
if (other.last_deep_scrub > last_deep_scrub) {
last_deep_scrub = other.last_deep_scrub;
modified = true;
}
if (other.last_deep_scrub_stamp > last_deep_scrub_stamp) {
last_deep_scrub_stamp = other.last_deep_scrub_stamp;
modified = true;
}
if (other.last_clean_scrub_stamp > last_clean_scrub_stamp) {
last_clean_scrub_stamp = other.last_clean_scrub_stamp;
modified = true;
}
return modified;
}
void encode(bufferlist& bl) const;
void decode(bufferlist::const_iterator& p);
void dump(Formatter *f) const;
static void generate_test_instances(list<pg_history_t*>& o);
};
WRITE_CLASS_ENCODER(pg_history_t)
inline ostream& operator<<(ostream& out, const pg_history_t& h) {
return out << "ec=" << h.epoch_created << "/" << h.epoch_pool_created
<< " lis/c " << h.last_interval_started
<< "/" << h.last_interval_clean
<< " les/c/f " << h.last_epoch_started << "/" << h.last_epoch_clean
<< "/" << h.last_epoch_marked_full
<< " " << h.same_up_since
<< "/" << h.same_interval_since
<< "/" << h.same_primary_since;
}
/**
* pg_info_t - summary of PG statistics.
*
* some notes:
* - last_complete implies we have all objects that existed as of that
* stamp, OR a newer object, OR have already applied a later delete.
* - if last_complete >= log.bottom, then we know pg contents thru log.head.
* otherwise, we have no idea what the pg is supposed to contain.
*/
struct pg_info_t {
spg_t pgid;
eversion_t last_update; ///< last object version applied to store.
eversion_t last_complete; ///< last version pg was complete through.
epoch_t last_epoch_started; ///< last epoch at which this pg started on this osd
epoch_t last_interval_started; ///< first epoch of last_epoch_started interval
version_t last_user_version; ///< last user object version applied to store
eversion_t log_tail; ///< oldest log entry.
hobject_t last_backfill; ///< objects >= this and < last_complete may be missing
bool last_backfill_bitwise; ///< true if last_backfill reflects a bitwise (vs nibblewise) sort
interval_set<snapid_t> purged_snaps;
pg_stat_t stats;
pg_history_t history;
pg_hit_set_history_t hit_set;
friend bool operator==(const pg_info_t& l, const pg_info_t& r) {
return
l.pgid == r.pgid &&
l.last_update == r.last_update &&
l.last_complete == r.last_complete &&
l.last_epoch_started == r.last_epoch_started &&
l.last_interval_started == r.last_interval_started &&
l.last_user_version == r.last_user_version &&
l.log_tail == r.log_tail &&
l.last_backfill == r.last_backfill &&
l.last_backfill_bitwise == r.last_backfill_bitwise &&
l.purged_snaps == r.purged_snaps &&
l.stats == r.stats &&
l.history == r.history &&
l.hit_set == r.hit_set;
}
pg_info_t()
: last_epoch_started(0),
last_interval_started(0),
last_user_version(0),
last_backfill(hobject_t::get_max()),
last_backfill_bitwise(false)
{ }
// cppcheck-suppress noExplicitConstructor
pg_info_t(spg_t p)
: pgid(p),
last_epoch_started(0),
last_interval_started(0),
last_user_version(0),
last_backfill(hobject_t::get_max()),
last_backfill_bitwise(false)
{ }
void set_last_backfill(hobject_t pos) {
last_backfill = pos;
last_backfill_bitwise = true;
}
bool is_empty() const { return last_update.version == 0; }
bool dne() const { return history.epoch_created == 0; }
bool has_missing() const { return last_complete != last_update; }
bool is_incomplete() const { return !last_backfill.is_max(); }
void encode(bufferlist& bl) const;
void decode(bufferlist::const_iterator& p);
void dump(Formatter *f) const;
static void generate_test_instances(list<pg_info_t*>& o);
};
WRITE_CLASS_ENCODER(pg_info_t)
inline ostream& operator<<(ostream& out, const pg_info_t& pgi)
{
out << pgi.pgid << "(";
if (pgi.dne())
out << " DNE";
if (pgi.is_empty())
out << " empty";
else {
out << " v " << pgi.last_update;
if (pgi.last_complete != pgi.last_update)
out << " lc " << pgi.last_complete;
out << " (" << pgi.log_tail << "," << pgi.last_update << "]";
}
if (pgi.is_incomplete())
out << " lb " << pgi.last_backfill
<< (pgi.last_backfill_bitwise ? " (bitwise)" : " (NIBBLEWISE)");
//out << " c " << pgi.epoch_created;
out << " local-lis/les=" << pgi.last_interval_started
<< "/" << pgi.last_epoch_started;
out << " n=" << pgi.stats.stats.sum.num_objects;
out << " " << pgi.history
<< ")";
return out;
}
/**
* pg_fast_info_t - common pg_info_t fields
*
* These are the fields of pg_info_t (and children) that are updated for
* most IO operations.
*
* ** WARNING **
* Because we rely on these fields to be applied to the normal
* info struct, adding a new field here that is not also new in info
* means that we must set an incompat OSD feature bit!
*/
struct pg_fast_info_t {
eversion_t last_update;
eversion_t last_complete;
version_t last_user_version;
struct { // pg_stat_t stats
eversion_t version;
version_t reported_seq;
utime_t last_fresh;
utime_t last_active;
utime_t last_peered;
utime_t last_clean;
utime_t last_unstale;
utime_t last_undegraded;
utime_t last_fullsized;
int64_t log_size; // (also ondisk_log_size, which has the same value)
struct { // object_stat_collection_t stats;
struct { // objct_stat_sum_t sum
int64_t num_bytes; // in bytes
int64_t num_objects;
int64_t num_object_copies;
int64_t num_rd;
int64_t num_rd_kb;
int64_t num_wr;
int64_t num_wr_kb;
int64_t num_objects_dirty;
} sum;
} stats;
} stats;
void populate_from(const pg_info_t& info) {
last_update = info.last_update;
last_complete = info.last_complete;
last_user_version = info.last_user_version;
stats.version = info.stats.version;
stats.reported_seq = info.stats.reported_seq;
stats.last_fresh = info.stats.last_fresh;
stats.last_active = info.stats.last_active;
stats.last_peered = info.stats.last_peered;
stats.last_clean = info.stats.last_clean;
stats.last_unstale = info.stats.last_unstale;
stats.last_undegraded = info.stats.last_undegraded;
stats.last_fullsized = info.stats.last_fullsized;
stats.log_size = info.stats.log_size;
stats.stats.sum.num_bytes = info.stats.stats.sum.num_bytes;
stats.stats.sum.num_objects = info.stats.stats.sum.num_objects;
stats.stats.sum.num_object_copies = info.stats.stats.sum.num_object_copies;
stats.stats.sum.num_rd = info.stats.stats.sum.num_rd;
stats.stats.sum.num_rd_kb = info.stats.stats.sum.num_rd_kb;
stats.stats.sum.num_wr = info.stats.stats.sum.num_wr;
stats.stats.sum.num_wr_kb = info.stats.stats.sum.num_wr_kb;
stats.stats.sum.num_objects_dirty = info.stats.stats.sum.num_objects_dirty;
}
bool try_apply_to(pg_info_t* info) {
if (last_update <= info->last_update)
return false;
info->last_update = last_update;
info->last_complete = last_complete;
info->last_user_version = last_user_version;
info->stats.version = stats.version;
info->stats.reported_seq = stats.reported_seq;
info->stats.last_fresh = stats.last_fresh;
info->stats.last_active = stats.last_active;
info->stats.last_peered = stats.last_peered;
info->stats.last_clean = stats.last_clean;
info->stats.last_unstale = stats.last_unstale;
info->stats.last_undegraded = stats.last_undegraded;
info->stats.last_fullsized = stats.last_fullsized;
info->stats.log_size = stats.log_size;
info->stats.ondisk_log_size = stats.log_size;
info->stats.stats.sum.num_bytes = stats.stats.sum.num_bytes;
info->stats.stats.sum.num_objects = stats.stats.sum.num_objects;
info->stats.stats.sum.num_object_copies = stats.stats.sum.num_object_copies;
info->stats.stats.sum.num_rd = stats.stats.sum.num_rd;
info->stats.stats.sum.num_rd_kb = stats.stats.sum.num_rd_kb;
info->stats.stats.sum.num_wr = stats.stats.sum.num_wr;
info->stats.stats.sum.num_wr_kb = stats.stats.sum.num_wr_kb;
info->stats.stats.sum.num_objects_dirty = stats.stats.sum.num_objects_dirty;
return true;
}
void encode(bufferlist& bl) const {
ENCODE_START(1, 1, bl);
encode(last_update, bl);
encode(last_complete, bl);
encode(last_user_version, bl);
encode(stats.version, bl);
encode(stats.reported_seq, bl);
encode(stats.last_fresh, bl);
encode(stats.last_active, bl);
encode(stats.last_peered, bl);
encode(stats.last_clean, bl);
encode(stats.last_unstale, bl);
encode(stats.last_undegraded, bl);
encode(stats.last_fullsized, bl);
encode(stats.log_size, bl);
encode(stats.stats.sum.num_bytes, bl);
encode(stats.stats.sum.num_objects, bl);
encode(stats.stats.sum.num_object_copies, bl);
encode(stats.stats.sum.num_rd, bl);
encode(stats.stats.sum.num_rd_kb, bl);
encode(stats.stats.sum.num_wr, bl);
encode(stats.stats.sum.num_wr_kb, bl);
encode(stats.stats.sum.num_objects_dirty, bl);
ENCODE_FINISH(bl);
}
void decode(bufferlist::const_iterator& p) {
DECODE_START(1, p);
decode(last_update, p);
decode(last_complete, p);
decode(last_user_version, p);
decode(stats.version, p);
decode(stats.reported_seq, p);
decode(stats.last_fresh, p);
decode(stats.last_active, p);
decode(stats.last_peered, p);
decode(stats.last_clean, p);
decode(stats.last_unstale, p);
decode(stats.last_undegraded, p);
decode(stats.last_fullsized, p);
decode(stats.log_size, p);
decode(stats.stats.sum.num_bytes, p);
decode(stats.stats.sum.num_objects, p);
decode(stats.stats.sum.num_object_copies, p);
decode(stats.stats.sum.num_rd, p);
decode(stats.stats.sum.num_rd_kb, p);
decode(stats.stats.sum.num_wr, p);
decode(stats.stats.sum.num_wr_kb, p);
decode(stats.stats.sum.num_objects_dirty, p);
DECODE_FINISH(p);
}
};
WRITE_CLASS_ENCODER(pg_fast_info_t)
struct pg_notify_t {
epoch_t query_epoch;
epoch_t epoch_sent;
pg_info_t info;
shard_id_t to;
shard_id_t from;
pg_notify_t() :
query_epoch(0), epoch_sent(0), to(shard_id_t::NO_SHARD),
from(shard_id_t::NO_SHARD) {}
pg_notify_t(
shard_id_t to,
shard_id_t from,
epoch_t query_epoch,
epoch_t epoch_sent,
const pg_info_t &info)
: query_epoch(query_epoch),
epoch_sent(epoch_sent),
info(info), to(to), from(from) {
ceph_assert(from == info.pgid.shard);
}
void encode(bufferlist &bl) const;
void decode(bufferlist::const_iterator &p);
void dump(Formatter *f) const;
static void generate_test_instances(list<pg_notify_t*> &o);
};
WRITE_CLASS_ENCODER(pg_notify_t)
ostream &operator<<(ostream &lhs, const pg_notify_t ¬ify);
class OSDMap;
/**
* PastIntervals -- information needed to determine the PriorSet and
* the might_have_unfound set
*/
class PastIntervals {
public:
struct pg_interval_t {
vector<int32_t> up, acting;
epoch_t first, last;
bool maybe_went_rw;
int32_t primary;
int32_t up_primary;
pg_interval_t()
: first(0), last(0),
maybe_went_rw(false),
primary(-1),
up_primary(-1)
{}
pg_interval_t(
vector<int32_t> &&up,
vector<int32_t> &&acting,
epoch_t first,
epoch_t last,
bool maybe_went_rw,
int32_t primary,
int32_t up_primary)
: up(up), acting(acting), first(first), last(last),
maybe_went_rw(maybe_went_rw), primary(primary), up_primary(up_primary)
{}
void encode(bufferlist& bl) const;
void decode(bufferlist::const_iterator& bl);
void dump(Formatter *f) const;
static void generate_test_instances(list<pg_interval_t*>& o);
};
PastIntervals();
PastIntervals(PastIntervals &&rhs) = default;
PastIntervals &operator=(PastIntervals &&rhs) = default;
PastIntervals(const PastIntervals &rhs);
PastIntervals &operator=(const PastIntervals &rhs);
class interval_rep {
public:
virtual size_t size() const = 0;
virtual bool empty() const = 0;
virtual void clear() = 0;
virtual pair<epoch_t, epoch_t> get_bounds() const = 0;
virtual set<pg_shard_t> get_all_participants(
bool ec_pool) const = 0;
virtual void add_interval(bool ec_pool, const pg_interval_t &interval) = 0;
virtual unique_ptr<interval_rep> clone() const = 0;
virtual ostream &print(ostream &out) const = 0;
virtual void encode(bufferlist &bl) const = 0;
virtual void decode(bufferlist::const_iterator &bl) = 0;
virtual void dump(Formatter *f) const = 0;
virtual void iterate_mayberw_back_to(
epoch_t les,
std::function<void(epoch_t, const set<pg_shard_t> &)> &&f) const = 0;
virtual bool has_full_intervals() const { return false; }
virtual void iterate_all_intervals(
std::function<void(const pg_interval_t &)> &&f) const {
ceph_assert(!has_full_intervals());
ceph_abort_msg("not valid for this implementation");
}
virtual void adjust_start_backwards(epoch_t last_epoch_clean) = 0;
virtual ~interval_rep() {}
};
friend class pi_compact_rep;
private:
unique_ptr<interval_rep> past_intervals;
explicit PastIntervals(interval_rep *rep) : past_intervals(rep) {}
public:
void add_interval(bool ec_pool, const pg_interval_t &interval) {
ceph_assert(past_intervals);
return past_intervals->add_interval(ec_pool, interval);
}
void encode(bufferlist &bl) const {
ENCODE_START(1, 1, bl);
if (past_intervals) {
__u8 type = 2;
encode(type, bl);
past_intervals->encode(bl);
} else {
encode((__u8)0, bl);
}
ENCODE_FINISH(bl);
}
void decode(bufferlist::const_iterator &bl);
void dump(Formatter *f) const {
ceph_assert(past_intervals);
past_intervals->dump(f);
}
static void generate_test_instances(list<PastIntervals *> & o);
/**
* Determines whether there is an interval change
*/
static bool is_new_interval(
int old_acting_primary,
int new_acting_primary,
const vector<int> &old_acting,
const vector<int> &new_acting,
int old_up_primary,
int new_up_primary,
const vector<int> &old_up,
const vector<int> &new_up,
int old_size,
int new_size,
int old_min_size,
int new_min_size,
unsigned old_pg_num,
unsigned new_pg_num,
unsigned old_pg_num_pending,
unsigned new_pg_num_pending,
bool old_sort_bitwise,
bool new_sort_bitwise,
bool old_recovery_deletes,
bool new_recovery_deletes,
pg_t pgid
);
/**
* Determines whether there is an interval change
*/
static bool is_new_interval(
int old_acting_primary, ///< [in] primary as of lastmap
int new_acting_primary, ///< [in] primary as of lastmap
const vector<int> &old_acting, ///< [in] acting as of lastmap
const vector<int> &new_acting, ///< [in] acting as of osdmap
int old_up_primary, ///< [in] up primary of lastmap
int new_up_primary, ///< [in] up primary of osdmap
const vector<int> &old_up, ///< [in] up as of lastmap
const vector<int> &new_up, ///< [in] up as of osdmap
std::shared_ptr<const OSDMap> osdmap, ///< [in] current map
std::shared_ptr<const OSDMap> lastmap, ///< [in] last map
pg_t pgid ///< [in] pgid for pg
);
/**
* Integrates a new map into *past_intervals, returns true
* if an interval was closed out.
*/
static bool check_new_interval(
int old_acting_primary, ///< [in] primary as of lastmap
int new_acting_primary, ///< [in] primary as of osdmap
const vector<int> &old_acting, ///< [in] acting as of lastmap
const vector<int> &new_acting, ///< [in] acting as of osdmap
int old_up_primary, ///< [in] up primary of lastmap
int new_up_primary, ///< [in] up primary of osdmap
const vector<int> &old_up, ///< [in] up as of lastmap
const vector<int> &new_up, ///< [in] up as of osdmap
epoch_t same_interval_since, ///< [in] as of osdmap
epoch_t last_epoch_clean, ///< [in] current
std::shared_ptr<const OSDMap> osdmap, ///< [in] current map
std::shared_ptr<const OSDMap> lastmap, ///< [in] last map
pg_t pgid, ///< [in] pgid for pg
IsPGRecoverablePredicate *could_have_gone_active, ///< [in] predicate whether the pg can be active
PastIntervals *past_intervals, ///< [out] intervals
ostream *out = 0 ///< [out] debug ostream
);
friend ostream& operator<<(ostream& out, const PastIntervals &i);
template <typename F>
void iterate_mayberw_back_to(
epoch_t les,
F &&f) const {
ceph_assert(past_intervals);
past_intervals->iterate_mayberw_back_to(les, std::forward<F>(f));
}
void clear() {
ceph_assert(past_intervals);
past_intervals->clear();
}
/**
* Should return a value which gives an indication of the amount
* of state contained
*/
size_t size() const {
ceph_assert(past_intervals);
return past_intervals->size();
}
bool empty() const {
ceph_assert(past_intervals);
return past_intervals->empty();
}
void swap(PastIntervals &other) {
using std::swap;
swap(other.past_intervals, past_intervals);
}
/**
* Return all shards which have been in the acting set back to the
* latest epoch to which we have trimmed except for pg_whoami
*/
set<pg_shard_t> get_might_have_unfound(
pg_shard_t pg_whoami,
bool ec_pool) const {
ceph_assert(past_intervals);
auto ret = past_intervals->get_all_participants(ec_pool);
ret.erase(pg_whoami);
return ret;
}
/**
* Return all shards which we might want to talk to for peering
*/
set<pg_shard_t> get_all_probe(
bool ec_pool) const {
ceph_assert(past_intervals);
return past_intervals->get_all_participants(ec_pool);
}
/* Return the set of epochs [start, end) represented by the
* past_interval set.
*/
pair<epoch_t, epoch_t> get_bounds() const {
ceph_assert(past_intervals);
return past_intervals->get_bounds();
}
void adjust_start_backwards(epoch_t last_epoch_clean) {
ceph_assert(past_intervals);
past_intervals->adjust_start_backwards(last_epoch_clean);
}
enum osd_state_t {
UP,
DOWN,
DNE,
LOST
};
struct PriorSet {
bool ec_pool = false;
set<pg_shard_t> probe; ///< current+prior OSDs we need to probe.
set<int> down; ///< down osds that would normally be in @a probe and might be interesting.
map<int, epoch_t> blocked_by; ///< current lost_at values for any OSDs in cur set for which (re)marking them lost would affect cur set
bool pg_down = false; ///< some down osds are included in @a cur; the DOWN pg state bit should be set.
unique_ptr<IsPGRecoverablePredicate> pcontdec;
PriorSet() = default;
PriorSet(PriorSet &&) = default;
PriorSet &operator=(PriorSet &&) = default;
PriorSet &operator=(const PriorSet &) = delete;
PriorSet(const PriorSet &) = delete;
bool operator==(const PriorSet &rhs) const {
return (ec_pool == rhs.ec_pool) &&
(probe == rhs.probe) &&
(down == rhs.down) &&
(blocked_by == rhs.blocked_by) &&
(pg_down == rhs.pg_down);
}
bool affected_by_map(
const OSDMap &osdmap,
const DoutPrefixProvider *dpp) const;
// For verifying tests
PriorSet(
bool ec_pool,
set<pg_shard_t> probe,
set<int> down,
map<int, epoch_t> blocked_by,
bool pg_down,
IsPGRecoverablePredicate *pcontdec)
: ec_pool(ec_pool), probe(probe), down(down), blocked_by(blocked_by),
pg_down(pg_down), pcontdec(pcontdec) {}
private:
template <typename F>
PriorSet(
const PastIntervals &past_intervals,
bool ec_pool,
epoch_t last_epoch_started,
IsPGRecoverablePredicate *c,
F f,
const vector<int> &up,
const vector<int> &acting,
const DoutPrefixProvider *dpp);
friend class PastIntervals;
};
template <typename... Args>
PriorSet get_prior_set(Args&&... args) const {
return PriorSet(*this, std::forward<Args>(args)...);
}
};
WRITE_CLASS_ENCODER(PastIntervals)
ostream& operator<<(ostream& out, const PastIntervals::pg_interval_t& i);
ostream& operator<<(ostream& out, const PastIntervals &i);
ostream& operator<<(ostream& out, const PastIntervals::PriorSet &i);
template <typename F>
PastIntervals::PriorSet::PriorSet(
const PastIntervals &past_intervals,
bool ec_pool,
epoch_t last_epoch_started,
IsPGRecoverablePredicate *c,
F f,
const vector<int> &up,
const vector<int> &acting,
const DoutPrefixProvider *dpp)
: ec_pool(ec_pool), pg_down(false), pcontdec(c)
{
/*
* We have to be careful to gracefully deal with situations like
* so. Say we have a power outage or something that takes out both
* OSDs, but the monitor doesn't mark them down in the same epoch.
* The history may look like
*
* 1: A B
* 2: B
* 3: let's say B dies for good, too (say, from the power spike)
* 4: A
*
* which makes it look like B may have applied updates to the PG
* that we need in order to proceed. This sucks...
*
* To minimize the risk of this happening, we CANNOT go active if
* _any_ OSDs in the prior set are down until we send an MOSDAlive
* to the monitor such that the OSDMap sets osd_up_thru to an epoch.
* Then, we have something like
*
* 1: A B
* 2: B up_thru[B]=0
* 3:
* 4: A
*
* -> we can ignore B, bc it couldn't have gone active (alive_thru
* still 0).
*
* or,
*
* 1: A B
* 2: B up_thru[B]=0
* 3: B up_thru[B]=2
* 4:
* 5: A
*
* -> we must wait for B, bc it was alive through 2, and could have
* written to the pg.
*
* If B is really dead, then an administrator will need to manually
* intervene by marking the OSD as "lost."
*/
// Include current acting and up nodes... not because they may
// contain old data (this interval hasn't gone active, obviously),
// but because we want their pg_info to inform choose_acting(), and
// so that we know what they do/do not have explicitly before
// sending them any new info/logs/whatever.
for (unsigned i = 0; i < acting.size(); i++) {
if (acting[i] != 0x7fffffff /* CRUSH_ITEM_NONE, can't import crush.h here */)
probe.insert(pg_shard_t(acting[i], ec_pool ? shard_id_t(i) : shard_id_t::NO_SHARD));
}
// It may be possible to exclude the up nodes, but let's keep them in
// there for now.
for (unsigned i = 0; i < up.size(); i++) {
if (up[i] != 0x7fffffff /* CRUSH_ITEM_NONE, can't import crush.h here */)
probe.insert(pg_shard_t(up[i], ec_pool ? shard_id_t(i) : shard_id_t::NO_SHARD));
}
set<pg_shard_t> all_probe = past_intervals.get_all_probe(ec_pool);
ldpp_dout(dpp, 10) << "build_prior all_probe " << all_probe << dendl;
for (auto &&i: all_probe) {
switch (f(0, i.osd, nullptr)) {
case UP: {
probe.insert(i);
break;
}
case DNE:
case LOST:
case DOWN: {
down.insert(i.osd);
break;
}
}
}
past_intervals.iterate_mayberw_back_to(
last_epoch_started,
[&](epoch_t start, const set<pg_shard_t> &acting) {
ldpp_dout(dpp, 10) << "build_prior maybe_rw interval:" << start
<< ", acting: " << acting << dendl;
// look at candidate osds during this interval. each falls into
// one of three categories: up, down (but potentially
// interesting), or lost (down, but we won't wait for it).
set<pg_shard_t> up_now;
map<int, epoch_t> candidate_blocked_by;
// any candidates down now (that might have useful data)
bool any_down_now = false;
// consider ACTING osds
for (auto &&so: acting) {
epoch_t lost_at = 0;
switch (f(start, so.osd, &lost_at)) {
case UP: {
// include past acting osds if they are up.
up_now.insert(so);
break;
}
case DNE: {
ldpp_dout(dpp, 10) << "build_prior prior osd." << so.osd
<< " no longer exists" << dendl;
break;
}
case LOST: {
ldpp_dout(dpp, 10) << "build_prior prior osd." << so.osd
<< " is down, but lost_at " << lost_at << dendl;
up_now.insert(so);
break;
}
case DOWN: {
ldpp_dout(dpp, 10) << "build_prior prior osd." << so.osd
<< " is down" << dendl;
candidate_blocked_by[so.osd] = lost_at;
any_down_now = true;
break;
}
}
}
// if not enough osds survived this interval, and we may have gone rw,
// then we need to wait for one of those osds to recover to
// ensure that we haven't lost any information.
if (!(*pcontdec)(up_now) && any_down_now) {
// fixme: how do we identify a "clean" shutdown anyway?
ldpp_dout(dpp, 10) << "build_prior possibly went active+rw,"
<< " insufficient up; including down osds" << dendl;
ceph_assert(!candidate_blocked_by.empty());
pg_down = true;
blocked_by.insert(
candidate_blocked_by.begin(),
candidate_blocked_by.end());
}
});
ldpp_dout(dpp, 10) << "build_prior final: probe " << probe
<< " down " << down
<< " blocked_by " << blocked_by
<< (pg_down ? " pg_down":"")
<< dendl;
}
/**
* pg_query_t - used to ask a peer for information about a pg.
*
* note: if version=0, type=LOG, then we just provide our full log.
*/
struct pg_query_t {
enum {
INFO = 0,
LOG = 1,
MISSING = 4,
FULLLOG = 5,
};
std::string_view get_type_name() const {
switch (type) {
case INFO: return "info";
case LOG: return "log";
case MISSING: return "missing";
case FULLLOG: return "fulllog";
default: return "???";
}
}
__s32 type;
eversion_t since;
pg_history_t history;
epoch_t epoch_sent;
shard_id_t to;
shard_id_t from;
pg_query_t() : type(-1), epoch_sent(0), to(shard_id_t::NO_SHARD),
from(shard_id_t::NO_SHARD) {}
pg_query_t(
int t,
shard_id_t to,
shard_id_t from,
const pg_history_t& h,
epoch_t epoch_sent)
: type(t),
history(h),
epoch_sent(epoch_sent),
to(to), from(from) {
ceph_assert(t != LOG);
}
pg_query_t(
int t,
shard_id_t to,
shard_id_t from,
eversion_t s,
const pg_history_t& h,
epoch_t epoch_sent)
: type(t), since(s), history(h),
epoch_sent(epoch_sent), to(to), from(from) {
ceph_assert(t == LOG);
}
void encode(bufferlist &bl, uint64_t features) const;
void decode(bufferlist::const_iterator &bl);
void dump(Formatter *f) const;
static void generate_test_instances(list<pg_query_t*>& o);
};
WRITE_CLASS_ENCODER_FEATURES(pg_query_t)
inline ostream& operator<<(ostream& out, const pg_query_t& q) {
out << "query(" << q.get_type_name() << " " << q.since;
if (q.type == pg_query_t::LOG)
out << " " << q.history;
out << " epoch_sent " << q.epoch_sent;
out << ")";
return out;
}
class PGBackend;
class ObjectModDesc {
bool can_local_rollback;
bool rollback_info_completed;
// version required to decode, reflected in encode/decode version
__u8 max_required_version = 1;
public:
class Visitor {
public:
virtual void append(uint64_t old_offset) {}
virtual void setattrs(map<string, boost::optional<bufferlist> > &attrs) {}
virtual void rmobject(version_t old_version) {}
/**
* Used to support the unfound_lost_delete log event: if the stashed
* version exists, we unstash it, otherwise, we do nothing. This way
* each replica rolls back to whatever state it had prior to the attempt
* at mark unfound lost delete
*/
virtual void try_rmobject(version_t old_version) {
rmobject(old_version);
}
virtual void create() {}
virtual void update_snaps(const set<snapid_t> &old_snaps) {}
virtual void rollback_extents(
version_t gen,
const vector<pair<uint64_t, uint64_t> > &extents) {}
virtual ~Visitor() {}
};
void visit(Visitor *visitor) const;
mutable bufferlist bl;
enum ModID {
APPEND = 1,
SETATTRS = 2,
DELETE = 3,
CREATE = 4,
UPDATE_SNAPS = 5,
TRY_DELETE = 6,
ROLLBACK_EXTENTS = 7
};
ObjectModDesc() : can_local_rollback(true), rollback_info_completed(false) {
bl.reassign_to_mempool(mempool::mempool_osd_pglog);
}
void claim(ObjectModDesc &other) {
bl.clear();
bl.claim(other.bl);
can_local_rollback = other.can_local_rollback;
rollback_info_completed = other.rollback_info_completed;
}
void claim_append(ObjectModDesc &other) {
if (!can_local_rollback || rollback_info_completed)
return;
if (!other.can_local_rollback) {
mark_unrollbackable();
return;
}
bl.claim_append(other.bl);
rollback_info_completed = other.rollback_info_completed;
}
void swap(ObjectModDesc &other) {
bl.swap(other.bl);
using std::swap;
swap(other.can_local_rollback, can_local_rollback);
swap(other.rollback_info_completed, rollback_info_completed);
swap(other.max_required_version, max_required_version);
}
void append_id(ModID id) {
using ceph::encode;
uint8_t _id(id);
encode(_id, bl);
}
void append(uint64_t old_size) {
if (!can_local_rollback || rollback_info_completed)
return;
ENCODE_START(1, 1, bl);
append_id(APPEND);
encode(old_size, bl);
ENCODE_FINISH(bl);
}
void setattrs(map<string, boost::optional<bufferlist> > &old_attrs) {
if (!can_local_rollback || rollback_info_completed)
return;
ENCODE_START(1, 1, bl);
append_id(SETATTRS);
encode(old_attrs, bl);
ENCODE_FINISH(bl);
}
bool rmobject(version_t deletion_version) {
if (!can_local_rollback || rollback_info_completed)
return false;
ENCODE_START(1, 1, bl);
append_id(DELETE);
encode(deletion_version, bl);
ENCODE_FINISH(bl);
rollback_info_completed = true;
return true;
}
bool try_rmobject(version_t deletion_version) {
if (!can_local_rollback || rollback_info_completed)
return false;
ENCODE_START(1, 1, bl);
append_id(TRY_DELETE);
encode(deletion_version, bl);
ENCODE_FINISH(bl);
rollback_info_completed = true;
return true;
}
void create() {
if (!can_local_rollback || rollback_info_completed)
return;
rollback_info_completed = true;
ENCODE_START(1, 1, bl);
append_id(CREATE);
ENCODE_FINISH(bl);
}
void update_snaps(const set<snapid_t> &old_snaps) {
if (!can_local_rollback || rollback_info_completed)
return;
ENCODE_START(1, 1, bl);
append_id(UPDATE_SNAPS);
encode(old_snaps, bl);
ENCODE_FINISH(bl);
}
void rollback_extents(
version_t gen, const vector<pair<uint64_t, uint64_t> > &extents) {
ceph_assert(can_local_rollback);
ceph_assert(!rollback_info_completed);
if (max_required_version < 2)
max_required_version = 2;
ENCODE_START(2, 2, bl);
append_id(ROLLBACK_EXTENTS);
encode(gen, bl);
encode(extents, bl);
ENCODE_FINISH(bl);
}
// cannot be rolled back
void mark_unrollbackable() {
can_local_rollback = false;
bl.clear();
}
bool can_rollback() const {
return can_local_rollback;
}
bool empty() const {
return can_local_rollback && (bl.length() == 0);
}
bool requires_kraken() const {
return max_required_version >= 2;
}
/**
* Create fresh copy of bl bytes to avoid keeping large buffers around
* in the case that bl contains ptrs which point into a much larger
* message buffer
*/
void trim_bl() const {
if (bl.length() > 0)
bl.rebuild();
}
void encode(bufferlist &bl) const;
void decode(bufferlist::const_iterator &bl);
void dump(Formatter *f) const;
static void generate_test_instances(list<ObjectModDesc*>& o);
};
WRITE_CLASS_ENCODER(ObjectModDesc)
/**
* pg_log_entry_t - single entry/event in pg log
*
*/
struct pg_log_entry_t {
enum {
MODIFY = 1, // some unspecified modification (but not *all* modifications)
CLONE = 2, // cloned object from head
DELETE = 3, // deleted object
//BACKLOG = 4, // event invented by generate_backlog [obsolete]
LOST_REVERT = 5, // lost new version, revert to an older version.
LOST_DELETE = 6, // lost new version, revert to no object (deleted).
LOST_MARK = 7, // lost new version, now EIO
PROMOTE = 8, // promoted object from another tier
CLEAN = 9, // mark an object clean
ERROR = 10, // write that returned an error
};
static const char *get_op_name(int op) {
switch (op) {
case MODIFY:
return "modify";
case PROMOTE:
return "promote";
case CLONE:
return "clone";
case DELETE:
return "delete";
case LOST_REVERT:
return "l_revert";
case LOST_DELETE:
return "l_delete";
case LOST_MARK:
return "l_mark";
case CLEAN:
return "clean";
case ERROR:
return "error";
default:
return "unknown";
}
}
const char *get_op_name() const {
return get_op_name(op);
}
// describes state for a locally-rollbackable entry
ObjectModDesc mod_desc;
bufferlist snaps; // only for clone entries
hobject_t soid;
osd_reqid_t reqid; // caller+tid to uniquely identify request
mempool::osd_pglog::vector<pair<osd_reqid_t, version_t> > extra_reqids;
/// map extra_reqids by index to error return code (if any)
mempool::osd_pglog::map<uint32_t, int> extra_reqid_return_codes;
eversion_t version, prior_version, reverting_to;
version_t user_version; // the user version for this entry
utime_t mtime; // this is the _user_ mtime, mind you
int32_t return_code; // only stored for ERRORs for dup detection
__s32 op;
bool invalid_hash; // only when decoding sobject_t based entries
bool invalid_pool; // only when decoding pool-less hobject based entries
pg_log_entry_t()
: user_version(0), return_code(0), op(0),
invalid_hash(false), invalid_pool(false) {
snaps.reassign_to_mempool(mempool::mempool_osd_pglog);
}
pg_log_entry_t(int _op, const hobject_t& _soid,
const eversion_t& v, const eversion_t& pv,
version_t uv,
const osd_reqid_t& rid, const utime_t& mt,
int return_code)
: soid(_soid), reqid(rid), version(v), prior_version(pv), user_version(uv),
mtime(mt), return_code(return_code), op(_op),
invalid_hash(false), invalid_pool(false) {
snaps.reassign_to_mempool(mempool::mempool_osd_pglog);
}
bool is_clone() const { return op == CLONE; }
bool is_modify() const { return op == MODIFY; }
bool is_promote() const { return op == PROMOTE; }
bool is_clean() const { return op == CLEAN; }
bool is_lost_revert() const { return op == LOST_REVERT; }
bool is_lost_delete() const { return op == LOST_DELETE; }
bool is_lost_mark() const { return op == LOST_MARK; }
bool is_error() const { return op == ERROR; }
bool is_update() const {
return
is_clone() || is_modify() || is_promote() || is_clean() ||
is_lost_revert() || is_lost_mark();
}
bool is_delete() const {
return op == DELETE || op == LOST_DELETE;
}
bool can_rollback() const {
return mod_desc.can_rollback();
}
void mark_unrollbackable() {
mod_desc.mark_unrollbackable();
}
bool requires_kraken() const {
return mod_desc.requires_kraken();
}
// Errors are only used for dup detection, whereas
// the index by objects is used by recovery, copy_get,
// and other facilities that don't expect or need to
// be aware of error entries.
bool object_is_indexed() const {
return !is_error();
}
bool reqid_is_indexed() const {
return reqid != osd_reqid_t() &&
(op == MODIFY || op == DELETE || op == ERROR);
}
string get_key_name() const;
void encode_with_checksum(bufferlist& bl) const;
void decode_with_checksum(bufferlist::const_iterator& p);
void encode(bufferlist &bl) const;
void decode(bufferlist::const_iterator &bl);
void dump(Formatter *f) const;
static void generate_test_instances(list<pg_log_entry_t*>& o);
};
WRITE_CLASS_ENCODER(pg_log_entry_t)
ostream& operator<<(ostream& out, const pg_log_entry_t& e);
struct pg_log_dup_t {
osd_reqid_t reqid; // caller+tid to uniquely identify request
eversion_t version;
version_t user_version; // the user version for this entry
int32_t return_code; // only stored for ERRORs for dup detection
pg_log_dup_t()
: user_version(0), return_code(0)
{}
explicit pg_log_dup_t(const pg_log_entry_t& entry)
: reqid(entry.reqid), version(entry.version),
user_version(entry.user_version), return_code(entry.return_code)
{}
pg_log_dup_t(const eversion_t& v, version_t uv,
const osd_reqid_t& rid, int return_code)
: reqid(rid), version(v), user_version(uv),
return_code(return_code)
{}
string get_key_name() const;
void encode(bufferlist &bl) const;
void decode(bufferlist::const_iterator &bl);
void dump(Formatter *f) const;
static void generate_test_instances(list<pg_log_dup_t*>& o);
bool operator==(const pg_log_dup_t &rhs) const {
return reqid == rhs.reqid &&
version == rhs.version &&
user_version == rhs.user_version &&
return_code == rhs.return_code;
}
bool operator!=(const pg_log_dup_t &rhs) const {
return !(*this == rhs);
}
friend std::ostream& operator<<(std::ostream& out, const pg_log_dup_t& e);
};
WRITE_CLASS_ENCODER(pg_log_dup_t)
std::ostream& operator<<(std::ostream& out, const pg_log_dup_t& e);
/**
* pg_log_t - incremental log of recent pg changes.
*
* serves as a recovery queue for recent changes.
*/
struct pg_log_t {
/*
* head - newest entry (update|delete)
* tail - entry previous to oldest (update|delete) for which we have
* complete negative information.
* i.e. we can infer pg contents for any store whose last_update >= tail.
*/
eversion_t head; // newest entry
eversion_t tail; // version prior to oldest
protected:
// We can rollback rollback-able entries > can_rollback_to
eversion_t can_rollback_to;
// always <= can_rollback_to, indicates how far stashed rollback
// data can be found
eversion_t rollback_info_trimmed_to;
public:
// the actual log
mempool::osd_pglog::list<pg_log_entry_t> log;
// entries just for dup op detection ordered oldest to newest
mempool::osd_pglog::list<pg_log_dup_t> dups;
pg_log_t() = default;
pg_log_t(const eversion_t &last_update,
const eversion_t &log_tail,
const eversion_t &can_rollback_to,
const eversion_t &rollback_info_trimmed_to,
mempool::osd_pglog::list<pg_log_entry_t> &&entries,
mempool::osd_pglog::list<pg_log_dup_t> &&dup_entries)
: head(last_update), tail(log_tail), can_rollback_to(can_rollback_to),
rollback_info_trimmed_to(rollback_info_trimmed_to),
log(std::move(entries)), dups(std::move(dup_entries)) {}
pg_log_t(const eversion_t &last_update,
const eversion_t &log_tail,
const eversion_t &can_rollback_to,
const eversion_t &rollback_info_trimmed_to,
const std::list<pg_log_entry_t> &entries,
const std::list<pg_log_dup_t> &dup_entries)
: head(last_update), tail(log_tail), can_rollback_to(can_rollback_to),
rollback_info_trimmed_to(rollback_info_trimmed_to) {
for (auto &&entry: entries) {
log.push_back(entry);
}
for (auto &&entry: dup_entries) {
dups.push_back(entry);
}
}
void clear() {
eversion_t z;
rollback_info_trimmed_to = can_rollback_to = head = tail = z;
log.clear();
dups.clear();
}
eversion_t get_rollback_info_trimmed_to() const {
return rollback_info_trimmed_to;
}
eversion_t get_can_rollback_to() const {
return can_rollback_to;
}
pg_log_t split_out_child(pg_t child_pgid, unsigned split_bits) {
mempool::osd_pglog::list<pg_log_entry_t> oldlog, childlog;
oldlog.swap(log);
eversion_t old_tail;
unsigned mask = ~((~0)<<split_bits);
for (auto i = oldlog.begin();
i != oldlog.end();
) {
if ((i->soid.get_hash() & mask) == child_pgid.m_seed) {
childlog.push_back(*i);
} else {
log.push_back(*i);
}
oldlog.erase(i++);
}
// osd_reqid is unique, so it doesn't matter if there are extra
// dup entries in each pg. To avoid storing oid with the dup
// entries, just copy the whole list.
auto childdups(dups);
return pg_log_t(
head,
tail,
can_rollback_to,
rollback_info_trimmed_to,
std::move(childlog),
std::move(childdups));
}
mempool::osd_pglog::list<pg_log_entry_t> rewind_from_head(eversion_t newhead) {
ceph_assert(newhead >= tail);
mempool::osd_pglog::list<pg_log_entry_t>::iterator p = log.end();
mempool::osd_pglog::list<pg_log_entry_t> divergent;
while (true) {
if (p == log.begin()) {
// yikes, the whole thing is divergent!
using std::swap;
swap(divergent, log);
break;
}
--p;
if (p->version.version <= newhead.version) {
/*
* look at eversion.version here. we want to avoid a situation like:
* our log: 100'10 (0'0) m 10000004d3a.00000000/head by client4225.1:18529
* new log: 122'10 (0'0) m 10000004d3a.00000000/head by client4225.1:18529
* lower_bound = 100'9
* i.e, same request, different version. If the eversion.version is > the
* lower_bound, we it is divergent.
*/
++p;
divergent.splice(divergent.begin(), log, p, log.end());
break;
}
ceph_assert(p->version > newhead);
}
head = newhead;
if (can_rollback_to > newhead)
can_rollback_to = newhead;
if (rollback_info_trimmed_to > newhead)
rollback_info_trimmed_to = newhead;
return divergent;
}
void merge_from(const vector<pg_log_t*>& slogs, eversion_t last_update) {
log.clear();
// sort and merge dups
multimap<eversion_t,pg_log_dup_t> sorted;
for (auto& d : dups) {
sorted.emplace(d.version, d);
}
for (auto l : slogs) {
for (auto& d : l->dups) {
sorted.emplace(d.version, d);
}
}
dups.clear();
for (auto& i : sorted) {
dups.push_back(i.second);
}
head = last_update;
tail = last_update;
can_rollback_to = last_update;
rollback_info_trimmed_to = last_update;
}
bool empty() const {
return log.empty();
}
bool null() const {
return head.version == 0 && head.epoch == 0;
}
size_t approx_size() const {
return head.version - tail.version;
}
static void filter_log(spg_t import_pgid, const OSDMap &curmap,
const string &hit_set_namespace, const pg_log_t &in,
pg_log_t &out, pg_log_t &reject);
/**
* copy entries from the tail of another pg_log_t
*
* @param other pg_log_t to copy from
* @param from copy entries after this version
*/
void copy_after(CephContext* cct, const pg_log_t &other, eversion_t from);
/**
* copy up to N entries
*
* @param other source log
* @param max max number of entries to copy
*/
void copy_up_to(CephContext* cct, const pg_log_t &other, int max);
ostream& print(ostream& out) const;
void encode(bufferlist &bl) const;
void decode(bufferlist::const_iterator &bl, int64_t pool = -1);
void dump(Formatter *f) const;
static void generate_test_instances(list<pg_log_t*>& o);
};
WRITE_CLASS_ENCODER(pg_log_t)
inline ostream& operator<<(ostream& out, const pg_log_t& log)
{
out << "log((" << log.tail << "," << log.head << "], crt="
<< log.get_can_rollback_to() << ")";
return out;
}
/**
* pg_missing_t - summary of missing objects.
*
* kept in memory, as a supplement to pg_log_t
* also used to pass missing info in messages.
*/
struct pg_missing_item {
eversion_t need, have;
enum missing_flags_t {
FLAG_NONE = 0,
FLAG_DELETE = 1,
} flags;
pg_missing_item() : flags(FLAG_NONE) {}
explicit pg_missing_item(eversion_t n) : need(n), flags(FLAG_NONE) {} // have no old version
pg_missing_item(eversion_t n, eversion_t h, bool is_delete=false) : need(n), have(h) {
set_delete(is_delete);
}
void encode(bufferlist& bl, uint64_t features) const {
using ceph::encode;
if (HAVE_FEATURE(features, OSD_RECOVERY_DELETES)) {
// encoding a zeroed eversion_t to differentiate between this and
// legacy unversioned encoding - a need value of 0'0 is not
// possible. This can be replaced with the legacy encoding
// macros post-luminous.
eversion_t e;
encode(e, bl);
encode(need, bl);
encode(have, bl);
encode(static_cast<uint8_t>(flags), bl);
} else {
// legacy unversioned encoding
encode(need, bl);
encode(have, bl);
}
}
void decode(bufferlist::const_iterator& bl) {
using ceph::decode;
eversion_t e;
decode(e, bl);
if (e != eversion_t()) {
// legacy encoding, this is the need value
need = e;
decode(have, bl);
} else {
decode(need, bl);
decode(have, bl);
uint8_t f;
decode(f, bl);
flags = static_cast<missing_flags_t>(f);
}
}
void set_delete(bool is_delete) {
flags = is_delete ? FLAG_DELETE : FLAG_NONE;
}
bool is_delete() const {
return (flags & FLAG_DELETE) == FLAG_DELETE;
}
string flag_str() const {
if (flags == FLAG_NONE) {
return "none";
} else {
return "delete";
}
}
void dump(Formatter *f) const {
f->dump_stream("need") << need;
f->dump_stream("have") << have;
f->dump_stream("flags") << flag_str();
}
static void generate_test_instances(list<pg_missing_item*>& o) {
o.push_back(new pg_missing_item);
o.push_back(new pg_missing_item);
o.back()->need = eversion_t(1, 2);
o.back()->have = eversion_t(1, 1);
o.push_back(new pg_missing_item);
o.back()->need = eversion_t(3, 5);
o.back()->have = eversion_t(3, 4);
o.back()->flags = FLAG_DELETE;
}
bool operator==(const pg_missing_item &rhs) const {
return need == rhs.need && have == rhs.have && flags == rhs.flags;
}
bool operator!=(const pg_missing_item &rhs) const {
return !(*this == rhs);
}
};
WRITE_CLASS_ENCODER_FEATURES(pg_missing_item)
ostream& operator<<(ostream& out, const pg_missing_item &item);
class pg_missing_const_i {
public:
virtual const map<hobject_t, pg_missing_item> &
get_items() const = 0;
virtual const map<version_t, hobject_t> &get_rmissing() const = 0;
virtual bool get_may_include_deletes() const = 0;
virtual unsigned int num_missing() const = 0;
virtual bool have_missing() const = 0;
virtual bool is_missing(const hobject_t& oid, pg_missing_item *out = nullptr) const = 0;
virtual bool is_missing(const hobject_t& oid, eversion_t v) const = 0;
virtual ~pg_missing_const_i() {}
};
template <bool Track>
class ChangeTracker {
public:
void changed(const hobject_t &obj) {}
template <typename F>
void get_changed(F &&f) const {}
void flush() {}
bool is_clean() const {
return true;
}
};
template <>
class ChangeTracker<true> {
set<hobject_t> _changed;
public:
void changed(const hobject_t &obj) {
_changed.insert(obj);
}
template <typename F>
void get_changed(F &&f) const {
for (auto const &i: _changed) {
f(i);
}
}
void flush() {
_changed.clear();
}
bool is_clean() const {
return _changed.empty();
}
};
template <bool TrackChanges>
class pg_missing_set : public pg_missing_const_i {
using item = pg_missing_item;
map<hobject_t, item> missing; // oid -> (need v, have v)
map<version_t, hobject_t> rmissing; // v -> oid
ChangeTracker<TrackChanges> tracker;
public:
pg_missing_set() = default;
template <typename missing_type>
pg_missing_set(const missing_type &m) {
missing = m.get_items();
rmissing = m.get_rmissing();
may_include_deletes = m.get_may_include_deletes();
for (auto &&i: missing)
tracker.changed(i.first);
}
bool may_include_deletes = false;
const map<hobject_t, item> &get_items() const override {
return missing;
}
const map<version_t, hobject_t> &get_rmissing() const override {
return rmissing;
}
bool get_may_include_deletes() const override {
return may_include_deletes;
}
unsigned int num_missing() const override {
return missing.size();
}
bool have_missing() const override {
return !missing.empty();
}
bool is_missing(const hobject_t& oid, pg_missing_item *out = nullptr) const override {
auto iter = missing.find(oid);
if (iter == missing.end())
return false;
if (out)
*out = iter->second;
return true;
}
bool is_missing(const hobject_t& oid, eversion_t v) const override {
map<hobject_t, item>::const_iterator m =
missing.find(oid);
if (m == missing.end())
return false;
const item &item(m->second);
if (item.need > v)
return false;
return true;
}
eversion_t get_oldest_need() const {
if (missing.empty()) {
return eversion_t();
}
auto it = missing.find(rmissing.begin()->second);
ceph_assert(it != missing.end());
return it->second.need;
}
void claim(pg_missing_set& o) {
static_assert(!TrackChanges, "Can't use claim with TrackChanges");
missing.swap(o.missing);
rmissing.swap(o.rmissing);
}
/*
* this needs to be called in log order as we extend the log. it
* assumes missing is accurate up through the previous log entry.
*/
void add_next_event(const pg_log_entry_t& e) {
map<hobject_t, item>::iterator missing_it;
missing_it = missing.find(e.soid);
bool is_missing_divergent_item = missing_it != missing.end();
if (e.prior_version == eversion_t() || e.is_clone()) {
// new object.
if (is_missing_divergent_item) { // use iterator
rmissing.erase((missing_it->second).need.version);
missing_it->second = item(e.version, eversion_t(), e.is_delete()); // .have = nil
} else // create new element in missing map
missing[e.soid] = item(e.version, eversion_t(), e.is_delete()); // .have = nil
} else if (is_missing_divergent_item) {
// already missing (prior).
rmissing.erase((missing_it->second).need.version);
(missing_it->second).need = e.version; // leave .have unchanged.
missing_it->second.set_delete(e.is_delete());
} else {
// not missing, we must have prior_version (if any)
ceph_assert(!is_missing_divergent_item);
missing[e.soid] = item(e.version, e.prior_version, e.is_delete());
}
rmissing[e.version.version] = e.soid;
tracker.changed(e.soid);
}
void revise_need(hobject_t oid, eversion_t need, bool is_delete) {
if (missing.count(oid)) {
rmissing.erase(missing[oid].need.version);
missing[oid].need = need; // no not adjust .have
missing[oid].set_delete(is_delete);
} else {
missing[oid] = item(need, eversion_t(), is_delete);
}
rmissing[need.version] = oid;
tracker.changed(oid);
}
void revise_have(hobject_t oid, eversion_t have) {
if (missing.count(oid)) {
tracker.changed(oid);
missing[oid].have = have;
}
}
void add(const hobject_t& oid, eversion_t need, eversion_t have,
bool is_delete) {
missing[oid] = item(need, have, is_delete);
rmissing[need.version] = oid;
tracker.changed(oid);
}
void rm(const hobject_t& oid, eversion_t v) {
std::map<hobject_t, item>::iterator p = missing.find(oid);
if (p != missing.end() && p->second.need <= v)
rm(p);
}
void rm(std::map<hobject_t, item>::const_iterator m) {
tracker.changed(m->first);
rmissing.erase(m->second.need.version);
missing.erase(m);
}
void got(const hobject_t& oid, eversion_t v) {
std::map<hobject_t, item>::iterator p = missing.find(oid);
ceph_assert(p != missing.end());
ceph_assert(p->second.need <= v || p->second.is_delete());
got(p);
}
void got(std::map<hobject_t, item>::const_iterator m) {
tracker.changed(m->first);
rmissing.erase(m->second.need.version);
missing.erase(m);
}
void split_into(
pg_t child_pgid,
unsigned split_bits,
pg_missing_set *omissing) {
omissing->may_include_deletes = may_include_deletes;
unsigned mask = ~((~0)<<split_bits);
for (map<hobject_t, item>::iterator i = missing.begin();
i != missing.end();
) {
if ((i->first.get_hash() & mask) == child_pgid.m_seed) {
omissing->add(i->first, i->second.need, i->second.have,
i->second.is_delete());
rm(i++);
} else {
++i;
}
}
}
void clear() {
for (auto const &i: missing)
tracker.changed(i.first);
missing.clear();
rmissing.clear();
}
void encode(bufferlist &bl) const {
ENCODE_START(4, 2, bl);
encode(missing, bl, may_include_deletes ? CEPH_FEATURE_OSD_RECOVERY_DELETES : 0);
encode(may_include_deletes, bl);
ENCODE_FINISH(bl);
}
void decode(bufferlist::const_iterator &bl, int64_t pool = -1) {
for (auto const &i: missing)
tracker.changed(i.first);
DECODE_START_LEGACY_COMPAT_LEN(4, 2, 2, bl);
decode(missing, bl);
if (struct_v >= 4) {
decode(may_include_deletes, bl);
}
DECODE_FINISH(bl);
if (struct_v < 3) {
// Handle hobject_t upgrade
map<hobject_t, item> tmp;
for (map<hobject_t, item>::iterator i =
missing.begin();
i != missing.end();
) {
if (!i->first.is_max() && i->first.pool == -1) {
hobject_t to_insert(i->first);
to_insert.pool = pool;
tmp[to_insert] = i->second;
missing.erase(i++);
} else {
++i;
}
}
missing.insert(tmp.begin(), tmp.end());
}
for (map<hobject_t,item>::iterator it =
missing.begin();
it != missing.end();
++it)
rmissing[it->second.need.version] = it->first;
for (auto const &i: missing)
tracker.changed(i.first);
}
void dump(Formatter *f) const {
f->open_array_section("missing");
for (map<hobject_t,item>::const_iterator p =
missing.begin(); p != missing.end(); ++p) {
f->open_object_section("item");
f->dump_stream("object") << p->first;
p->second.dump(f);
f->close_section();
}
f->close_section();
f->dump_bool("may_include_deletes", may_include_deletes);
}
template <typename F>
void filter_objects(F &&f) {
for (auto i = missing.begin(); i != missing.end();) {
if (f(i->first)) {
rm(i++);
} else {
++i;
}
}
}
static void generate_test_instances(list<pg_missing_set*>& o) {
o.push_back(new pg_missing_set);
o.push_back(new pg_missing_set);
o.back()->add(
hobject_t(object_t("foo"), "foo", 123, 456, 0, ""),
eversion_t(5, 6), eversion_t(5, 1), false);
o.push_back(new pg_missing_set);
o.back()->add(
hobject_t(object_t("foo"), "foo", 123, 456, 0, ""),
eversion_t(5, 6), eversion_t(5, 1), true);
o.back()->may_include_deletes = true;
}
template <typename F>
void get_changed(F &&f) const {
tracker.get_changed(f);
}
void flush() {
tracker.flush();
}
bool is_clean() const {
return tracker.is_clean();
}
template <typename missing_t>
bool debug_verify_from_init(
const missing_t &init_missing,
ostream *oss) const {
if (!TrackChanges)
return true;
auto check_missing(init_missing.get_items());
tracker.get_changed([&](const hobject_t &hoid) {
check_missing.erase(hoid);
if (missing.count(hoid)) {
check_missing.insert(*(missing.find(hoid)));
}
});
bool ok = true;
if (check_missing.size() != missing.size()) {
if (oss) {
*oss << "Size mismatch, check: " << check_missing.size()
<< ", actual: " << missing.size() << "\n";
}
ok = false;
}
for (auto &i: missing) {
if (!check_missing.count(i.first)) {
if (oss)
*oss << "check_missing missing " << i.first << "\n";
ok = false;
} else if (check_missing[i.first] != i.second) {
if (oss)
*oss << "check_missing missing item mismatch on " << i.first
<< ", check: " << check_missing[i.first]
<< ", actual: " << i.second << "\n";
ok = false;
}
}
if (oss && !ok) {
*oss << "check_missing: " << check_missing << "\n";
set<hobject_t> changed;
tracker.get_changed([&](const hobject_t &hoid) { changed.insert(hoid); });
*oss << "changed: " << changed << "\n";
}
return ok;
}
};
template <bool TrackChanges>
void encode(
const pg_missing_set<TrackChanges> &c, bufferlist &bl, uint64_t features=0) {
ENCODE_DUMP_PRE();
c.encode(bl);
ENCODE_DUMP_POST(cl);
}
template <bool TrackChanges>
void decode(pg_missing_set<TrackChanges> &c, bufferlist::const_iterator &p) {
c.decode(p);
}
template <bool TrackChanges>
ostream& operator<<(ostream& out, const pg_missing_set<TrackChanges> &missing)
{
out << "missing(" << missing.num_missing()
<< " may_include_deletes = " << missing.may_include_deletes;
//if (missing.num_lost()) out << ", " << missing.num_lost() << " lost";
out << ")";
return out;
}
using pg_missing_t = pg_missing_set<false>;
using pg_missing_tracker_t = pg_missing_set<true>;
/**
* pg list objects response format
*
*/
struct pg_nls_response_t {
collection_list_handle_t handle;
list<librados::ListObjectImpl> entries;
void encode(bufferlist& bl) const {
ENCODE_START(1, 1, bl);
encode(handle, bl);
__u32 n = (__u32)entries.size();
encode(n, bl);
for (list<librados::ListObjectImpl>::const_iterator i = entries.begin(); i != entries.end(); ++i) {
encode(i->nspace, bl);
encode(i->oid, bl);
encode(i->locator, bl);
}
ENCODE_FINISH(bl);
}
void decode(bufferlist::const_iterator& bl) {
DECODE_START(1, bl);
decode(handle, bl);
__u32 n;
decode(n, bl);
entries.clear();
while (n--) {
librados::ListObjectImpl i;
decode(i.nspace, bl);
decode(i.oid, bl);
decode(i.locator, bl);
entries.push_back(i);
}
DECODE_FINISH(bl);
}
void dump(Formatter *f) const {
f->dump_stream("handle") << handle;
f->open_array_section("entries");
for (list<librados::ListObjectImpl>::const_iterator p = entries.begin(); p != entries.end(); ++p) {
f->open_object_section("object");
f->dump_string("namespace", p->nspace);
f->dump_string("object", p->oid);
f->dump_string("key", p->locator);
f->close_section();
}
f->close_section();
}
static void generate_test_instances(list<pg_nls_response_t*>& o) {
o.push_back(new pg_nls_response_t);
o.push_back(new pg_nls_response_t);
o.back()->handle = hobject_t(object_t("hi"), "key", 1, 2, -1, "");
o.back()->entries.push_back(librados::ListObjectImpl("", "one", ""));
o.back()->entries.push_back(librados::ListObjectImpl("", "two", "twokey"));
o.back()->entries.push_back(librados::ListObjectImpl("", "three", ""));
o.push_back(new pg_nls_response_t);
o.back()->handle = hobject_t(object_t("hi"), "key", 3, 4, -1, "");
o.back()->entries.push_back(librados::ListObjectImpl("n1", "n1one", ""));
o.back()->entries.push_back(librados::ListObjectImpl("n1", "n1two", "n1twokey"));
o.back()->entries.push_back(librados::ListObjectImpl("n1", "n1three", ""));
o.push_back(new pg_nls_response_t);
o.back()->handle = hobject_t(object_t("hi"), "key", 5, 6, -1, "");
o.back()->entries.push_back(librados::ListObjectImpl("", "one", ""));
o.back()->entries.push_back(librados::ListObjectImpl("", "two", "twokey"));
o.back()->entries.push_back(librados::ListObjectImpl("", "three", ""));
o.back()->entries.push_back(librados::ListObjectImpl("n1", "n1one", ""));
o.back()->entries.push_back(librados::ListObjectImpl("n1", "n1two", "n1twokey"));
o.back()->entries.push_back(librados::ListObjectImpl("n1", "n1three", ""));
}
};
WRITE_CLASS_ENCODER(pg_nls_response_t)
// For backwards compatibility with older OSD requests
struct pg_ls_response_t {
collection_list_handle_t handle;
list<pair<object_t, string> > entries;
void encode(bufferlist& bl) const {
using ceph::encode;
__u8 v = 1;
encode(v, bl);
encode(handle, bl);
encode(entries, bl);
}
void decode(bufferlist::const_iterator& bl) {
using ceph::decode;
__u8 v;
decode(v, bl);
ceph_assert(v == 1);
decode(handle, bl);
decode(entries, bl);
}
void dump(Formatter *f) const {
f->dump_stream("handle") << handle;
f->open_array_section("entries");
for (list<pair<object_t, string> >::const_iterator p = entries.begin(); p != entries.end(); ++p) {
f->open_object_section("object");
f->dump_stream("object") << p->first;
f->dump_string("key", p->second);
f->close_section();
}
f->close_section();
}
static void generate_test_instances(list<pg_ls_response_t*>& o) {
o.push_back(new pg_ls_response_t);
o.push_back(new pg_ls_response_t);
o.back()->handle = hobject_t(object_t("hi"), "key", 1, 2, -1, "");
o.back()->entries.push_back(make_pair(object_t("one"), string()));
o.back()->entries.push_back(make_pair(object_t("two"), string("twokey")));
}
};
WRITE_CLASS_ENCODER(pg_ls_response_t)
/**
* object_copy_cursor_t
*/
struct object_copy_cursor_t {
uint64_t data_offset;
string omap_offset;
bool attr_complete;
bool data_complete;
bool omap_complete;
object_copy_cursor_t()
: data_offset(0),
attr_complete(false),
data_complete(false),
omap_complete(false)
{}
bool is_initial() const {
return !attr_complete && data_offset == 0 && omap_offset.empty();
}
bool is_complete() const {
return attr_complete && data_complete && omap_complete;
}
static void generate_test_instances(list<object_copy_cursor_t*>& o);
void encode(bufferlist& bl) const;
void decode(bufferlist::const_iterator &bl);
void dump(Formatter *f) const;
};
WRITE_CLASS_ENCODER(object_copy_cursor_t)
/**
* object_copy_data_t
*
* Return data from a copy request. The semantics are a little strange
* as a result of the encoding's heritage.
*
* In particular, the sender unconditionally fills in the cursor (from what
* it receives and sends), the size, and the mtime, but is responsible for
* figuring out whether it should put any data in the attrs, data, or
* omap members (corresponding to xattrs, object data, and the omap entries)
* based on external data (the client includes a max amount to return with
* the copy request). The client then looks into the attrs, data, and/or omap
* based on the contents of the cursor.
*/
struct object_copy_data_t {
enum {
FLAG_DATA_DIGEST = 1<<0,
FLAG_OMAP_DIGEST = 1<<1,
};
object_copy_cursor_t cursor;
uint64_t size;
utime_t mtime;
uint32_t data_digest, omap_digest;
uint32_t flags;
map<string, bufferlist> attrs;
bufferlist data;
bufferlist omap_header;
bufferlist omap_data;
/// which snaps we are defined for (if a snap and not the head)
vector<snapid_t> snaps;
/// latest snap seq for the object (if head)
snapid_t snap_seq;
/// recent reqids on this object
mempool::osd_pglog::vector<pair<osd_reqid_t, version_t> > reqids;
/// map reqids by index to error return code (if any)
mempool::osd_pglog::map<uint32_t, int> reqid_return_codes;
uint64_t truncate_seq;
uint64_t truncate_size;
public:
object_copy_data_t() :
size((uint64_t)-1), data_digest(-1),
omap_digest(-1), flags(0),
truncate_seq(0),
truncate_size(0) {}
static void generate_test_instances(list<object_copy_data_t*>& o);
void encode(bufferlist& bl, uint64_t features) const;
void decode(bufferlist::const_iterator& bl);
void dump(Formatter *f) const;
};
WRITE_CLASS_ENCODER_FEATURES(object_copy_data_t)
/**
* pg creation info
*/
struct pg_create_t {
epoch_t created; // epoch pg created
pg_t parent; // split from parent (if != pg_t())
__s32 split_bits;
pg_create_t()
: created(0), split_bits(0) {}
pg_create_t(unsigned c, pg_t p, int s)
: created(c), parent(p), split_bits(s) {}
void encode(bufferlist &bl) const;
void decode(bufferlist::const_iterator &bl);
void dump(Formatter *f) const;
static void generate_test_instances(list<pg_create_t*>& o);
};
WRITE_CLASS_ENCODER(pg_create_t)
// -----------------------------------------
class ObjectExtent {
/**
* ObjectExtents are used for specifying IO behavior against RADOS
* objects when one is using the ObjectCacher.
*
* To use this in a real system, *every member* must be filled
* out correctly. In particular, make sure to initialize the
* oloc correctly, as its default values are deliberate poison
* and will cause internal ObjectCacher asserts.
*
* Similarly, your buffer_extents vector *must* specify a total
* size equal to your length. If the buffer_extents inadvertently
* contain less space than the length member specifies, you
* will get unintelligible asserts deep in the ObjectCacher.
*
* If you are trying to do testing and don't care about actual
* RADOS function, the simplest thing to do is to initialize
* the ObjectExtent (truncate_size can be 0), create a single entry
* in buffer_extents matching the length, and set oloc.pool to 0.
*/
public:
object_t oid; // object id
uint64_t objectno;
uint64_t offset; // in object
uint64_t length; // in object
uint64_t truncate_size; // in object
object_locator_t oloc; // object locator (pool etc)
vector<pair<uint64_t,uint64_t> > buffer_extents; // off -> len. extents in buffer being mapped (may be fragmented bc of striping!)
ObjectExtent() : objectno(0), offset(0), length(0), truncate_size(0) {}
ObjectExtent(object_t o, uint64_t ono, uint64_t off, uint64_t l, uint64_t ts) :
oid(o), objectno(ono), offset(off), length(l), truncate_size(ts) { }
};
inline ostream& operator<<(ostream& out, const ObjectExtent &ex)
{
return out << "extent("
<< ex.oid << " (" << ex.objectno << ") in " << ex.oloc
<< " " << ex.offset << "~" << ex.length
<< " -> " << ex.buffer_extents
<< ")";
}
// ---------------------------------------
class OSDSuperblock {
public:
uuid_d cluster_fsid, osd_fsid;
int32_t whoami; // my role in this fs.
epoch_t current_epoch; // most recent epoch
epoch_t oldest_map, newest_map; // oldest/newest maps we have.
double weight;
CompatSet compat_features;
// last interval over which i mounted and was then active
epoch_t mounted; // last epoch i mounted
epoch_t clean_thru; // epoch i was active and clean thru
OSDSuperblock() :
whoami(-1),
current_epoch(0), oldest_map(0), newest_map(0), weight(0),
mounted(0), clean_thru(0) {
}
void encode(bufferlist &bl) const;
void decode(bufferlist::const_iterator &bl);
void dump(Formatter *f) const;
static void generate_test_instances(list<OSDSuperblock*>& o);
};
WRITE_CLASS_ENCODER(OSDSuperblock)
inline ostream& operator<<(ostream& out, const OSDSuperblock& sb)
{
return out << "sb(" << sb.cluster_fsid
<< " osd." << sb.whoami
<< " " << sb.osd_fsid
<< " e" << sb.current_epoch
<< " [" << sb.oldest_map << "," << sb.newest_map << "]"
<< " lci=[" << sb.mounted << "," << sb.clean_thru << "]"
<< ")";
}
// -------
/*
* attached to object head. describes most recent snap context, and
* set of existing clones.
*/
struct SnapSet {
snapid_t seq;
vector<snapid_t> snaps; // descending
vector<snapid_t> clones; // ascending
map<snapid_t, interval_set<uint64_t> > clone_overlap; // overlap w/ next newest
map<snapid_t, uint64_t> clone_size;
map<snapid_t, vector<snapid_t>> clone_snaps; // descending
SnapSet() : seq(0) {}
explicit SnapSet(bufferlist& bl) {
auto p = std::cbegin(bl);
decode(p);
}
/// populate SnapSet from a librados::snap_set_t
void from_snap_set(const librados::snap_set_t& ss, bool legacy);
/// get space accounted to clone
uint64_t get_clone_bytes(snapid_t clone) const;
void encode(bufferlist& bl) const;
void decode(bufferlist::const_iterator& bl);
void dump(Formatter *f) const;
static void generate_test_instances(list<SnapSet*>& o);
SnapContext get_ssc_as_of(snapid_t as_of) const {
SnapContext out;
out.seq = as_of;
for (vector<snapid_t>::const_iterator i = snaps.begin();
i != snaps.end();
++i) {
if (*i <= as_of)
out.snaps.push_back(*i);
}
return out;
}
SnapSet get_filtered(const pg_pool_t &pinfo) const;
void filter(const pg_pool_t &pinfo);
};
WRITE_CLASS_ENCODER(SnapSet)
ostream& operator<<(ostream& out, const SnapSet& cs);
#define OI_ATTR "_"
#define SS_ATTR "snapset"
struct watch_info_t {
uint64_t cookie;
uint32_t timeout_seconds;
entity_addr_t addr;
watch_info_t() : cookie(0), timeout_seconds(0) { }
watch_info_t(uint64_t c, uint32_t t, const entity_addr_t& a) : cookie(c), timeout_seconds(t), addr(a) {}
void encode(bufferlist& bl, uint64_t features) const;
void decode(bufferlist::const_iterator& bl);
void dump(Formatter *f) const;
static void generate_test_instances(list<watch_info_t*>& o);
};
WRITE_CLASS_ENCODER_FEATURES(watch_info_t)
static inline bool operator==(const watch_info_t& l, const watch_info_t& r) {
return l.cookie == r.cookie && l.timeout_seconds == r.timeout_seconds
&& l.addr == r.addr;
}
static inline ostream& operator<<(ostream& out, const watch_info_t& w) {
return out << "watch(cookie " << w.cookie << " " << w.timeout_seconds << "s"
<< " " << w.addr << ")";
}
struct notify_info_t {
uint64_t cookie;
uint64_t notify_id;
uint32_t timeout;
bufferlist bl;
};
static inline ostream& operator<<(ostream& out, const notify_info_t& n) {
return out << "notify(cookie " << n.cookie
<< " notify" << n.notify_id
<< " " << n.timeout << "s)";
}
struct chunk_info_t {
typedef enum {
FLAG_DIRTY = 1,
FLAG_MISSING = 2,
FLAG_HAS_REFERENCE = 4,
FLAG_HAS_FINGERPRINT = 8,
} cflag_t;
uint32_t offset;
uint32_t length;
hobject_t oid;
cflag_t flags; // FLAG_*
chunk_info_t() : offset(0), length(0), flags((cflag_t)0) { }
static string get_flag_string(uint64_t flags) {
string r;
if (flags & FLAG_DIRTY) {
r += "|dirty";
}
if (flags & FLAG_MISSING) {
r += "|missing";
}
if (flags & FLAG_HAS_REFERENCE) {
r += "|has_reference";
}
if (flags & FLAG_HAS_FINGERPRINT) {
r += "|has_fingerprint";
}
if (r.length())
return r.substr(1);
return r;
}
bool test_flag(cflag_t f) const {
return (flags & f) == f;
}
void set_flag(cflag_t f) {
flags = (cflag_t)(flags | f);
}
void set_flags(cflag_t f) {
flags = f;
}
void clear_flag(cflag_t f) {
flags = (cflag_t)(flags & ~f);
}
void clear_flags() {
flags = (cflag_t)0;
}
bool is_dirty() const {
return test_flag(FLAG_DIRTY);
}
bool is_missing() const {
return test_flag(FLAG_MISSING);
}
bool has_reference() const {
return test_flag(FLAG_HAS_REFERENCE);
}
bool has_fingerprint() const {
return test_flag(FLAG_HAS_FINGERPRINT);
}
void encode(bufferlist &bl) const;
void decode(bufferlist::const_iterator &bl);
void dump(Formatter *f) const;
friend ostream& operator<<(ostream& out, const chunk_info_t& ci);
};
WRITE_CLASS_ENCODER(chunk_info_t)
ostream& operator<<(ostream& out, const chunk_info_t& ci);
struct object_info_t;
struct object_manifest_t {
enum {
TYPE_NONE = 0,
TYPE_REDIRECT = 1,
TYPE_CHUNKED = 2,
};
uint8_t type; // redirect, chunked, ...
hobject_t redirect_target;
map <uint64_t, chunk_info_t> chunk_map;
object_manifest_t() : type(0) { }
object_manifest_t(uint8_t type, const hobject_t& redirect_target)
: type(type), redirect_target(redirect_target) { }
bool is_empty() const {
return type == TYPE_NONE;
}
bool is_redirect() const {
return type == TYPE_REDIRECT;
}
bool is_chunked() const {
return type == TYPE_CHUNKED;
}
static std::string_view get_type_name(uint8_t m) {
switch (m) {
case TYPE_NONE: return "none";
case TYPE_REDIRECT: return "redirect";
case TYPE_CHUNKED: return "chunked";
default: return "unknown";
}
}
std::string_view get_type_name() const {
return get_type_name(type);
}
void clear() {
type = 0;
redirect_target = hobject_t();
chunk_map.clear();
}
static void generate_test_instances(list<object_manifest_t*>& o);
void encode(bufferlist &bl) const;
void decode(bufferlist::const_iterator &bl);
void dump(Formatter *f) const;
friend ostream& operator<<(ostream& out, const object_info_t& oi);
};
WRITE_CLASS_ENCODER(object_manifest_t)
ostream& operator<<(ostream& out, const object_manifest_t& oi);
struct object_info_t {
hobject_t soid;
eversion_t version, prior_version;
version_t user_version;
osd_reqid_t last_reqid;
uint64_t size;
utime_t mtime;
utime_t local_mtime; // local mtime
// note: these are currently encoded into a total 16 bits; see
// encode()/decode() for the weirdness.
typedef enum {
FLAG_LOST = 1<<0,
FLAG_WHITEOUT = 1<<1, // object logically does not exist
FLAG_DIRTY = 1<<2, // object has been modified since last flushed or undirtied
FLAG_OMAP = 1<<3, // has (or may have) some/any omap data
FLAG_DATA_DIGEST = 1<<4, // has data crc
FLAG_OMAP_DIGEST = 1<<5, // has omap crc
FLAG_CACHE_PIN = 1<<6, // pin the object in cache tier
FLAG_MANIFEST = 1<<7, // has manifest
FLAG_USES_TMAP = 1<<8, // deprecated; no longer used
FLAG_REDIRECT_HAS_REFERENCE = 1<<9, // has reference
} flag_t;
flag_t flags;
static string get_flag_string(flag_t flags) {
string s;
vector<string> sv = get_flag_vector(flags);
for (auto ss : sv) {
s += string("|") + ss;
}
if (s.length())
return s.substr(1);
return s;
}
static vector<string> get_flag_vector(flag_t flags) {
vector<string> sv;
if (flags & FLAG_LOST)
sv.insert(sv.end(), "lost");
if (flags & FLAG_WHITEOUT)
sv.insert(sv.end(), "whiteout");
if (flags & FLAG_DIRTY)
sv.insert(sv.end(), "dirty");
if (flags & FLAG_USES_TMAP)
sv.insert(sv.end(), "uses_tmap");
if (flags & FLAG_OMAP)
sv.insert(sv.end(), "omap");
if (flags & FLAG_DATA_DIGEST)
sv.insert(sv.end(), "data_digest");
if (flags & FLAG_OMAP_DIGEST)
sv.insert(sv.end(), "omap_digest");
if (flags & FLAG_CACHE_PIN)
sv.insert(sv.end(), "cache_pin");
if (flags & FLAG_MANIFEST)
sv.insert(sv.end(), "manifest");
if (flags & FLAG_REDIRECT_HAS_REFERENCE)
sv.insert(sv.end(), "redirect_has_reference");
return sv;
}
string get_flag_string() const {
return get_flag_string(flags);
}
uint64_t truncate_seq, truncate_size;
map<pair<uint64_t, entity_name_t>, watch_info_t> watchers;
// opportunistic checksums; may or may not be present
__u32 data_digest; ///< data crc32c
__u32 omap_digest; ///< omap crc32c
// alloc hint attribute
uint64_t expected_object_size, expected_write_size;
uint32_t alloc_hint_flags;
struct object_manifest_t manifest;
void copy_user_bits(const object_info_t& other);
bool test_flag(flag_t f) const {
return (flags & f) == f;
}
void set_flag(flag_t f) {
flags = (flag_t)(flags | f);
}
void clear_flag(flag_t f) {
flags = (flag_t)(flags & ~f);
}
bool is_lost() const {
return test_flag(FLAG_LOST);
}
bool is_whiteout() const {
return test_flag(FLAG_WHITEOUT);
}
bool is_dirty() const {
return test_flag(FLAG_DIRTY);
}
bool is_omap() const {
return test_flag(FLAG_OMAP);
}
bool is_data_digest() const {
return test_flag(FLAG_DATA_DIGEST);
}
bool is_omap_digest() const {
return test_flag(FLAG_OMAP_DIGEST);
}
bool is_cache_pinned() const {
return test_flag(FLAG_CACHE_PIN);
}
bool has_manifest() const {
return test_flag(FLAG_MANIFEST);
}
void set_data_digest(__u32 d) {
set_flag(FLAG_DATA_DIGEST);
data_digest = d;
}
void set_omap_digest(__u32 d) {
set_flag(FLAG_OMAP_DIGEST);
omap_digest = d;
}
void clear_data_digest() {
clear_flag(FLAG_DATA_DIGEST);
data_digest = -1;
}
void clear_omap_digest() {
clear_flag(FLAG_OMAP_DIGEST);
omap_digest = -1;
}
void new_object() {
clear_data_digest();
clear_omap_digest();
}
void encode(bufferlist& bl, uint64_t features) const;
void decode(bufferlist::const_iterator& bl);
void decode(bufferlist& bl) {
auto p = std::cbegin(bl);
decode(p);
}
void dump(Formatter *f) const;
static void generate_test_instances(list<object_info_t*>& o);
explicit object_info_t()
: user_version(0), size(0), flags((flag_t)0),
truncate_seq(0), truncate_size(0),
data_digest(-1), omap_digest(-1),
expected_object_size(0), expected_write_size(0),
alloc_hint_flags(0)
{}
explicit object_info_t(const hobject_t& s)
: soid(s),
user_version(0), size(0), flags((flag_t)0),
truncate_seq(0), truncate_size(0),
data_digest(-1), omap_digest(-1),
expected_object_size(0), expected_write_size(0),
alloc_hint_flags(0)
{}
explicit object_info_t(bufferlist& bl) {
decode(bl);
}
};
WRITE_CLASS_ENCODER_FEATURES(object_info_t)
ostream& operator<<(ostream& out, const object_info_t& oi);
// Object recovery
struct ObjectRecoveryInfo {
hobject_t soid;
eversion_t version;
uint64_t size;
object_info_t oi;
SnapSet ss; // only populated if soid is_snap()
interval_set<uint64_t> copy_subset;
map<hobject_t, interval_set<uint64_t>> clone_subset;
ObjectRecoveryInfo() : size(0) { }
static void generate_test_instances(list<ObjectRecoveryInfo*>& o);
void encode(bufferlist &bl, uint64_t features) const;
void decode(bufferlist::const_iterator &bl, int64_t pool = -1);
ostream &print(ostream &out) const;
void dump(Formatter *f) const;
};
WRITE_CLASS_ENCODER_FEATURES(ObjectRecoveryInfo)
ostream& operator<<(ostream& out, const ObjectRecoveryInfo &inf);
struct ObjectRecoveryProgress {
uint64_t data_recovered_to;
string omap_recovered_to;
bool first;
bool data_complete;
bool omap_complete;
bool error = false;
ObjectRecoveryProgress()
: data_recovered_to(0),
first(true),
data_complete(false), omap_complete(false) { }
bool is_complete(const ObjectRecoveryInfo& info) const {
return (data_recovered_to >= (
info.copy_subset.empty() ?
0 : info.copy_subset.range_end())) &&
omap_complete;
}
static void generate_test_instances(list<ObjectRecoveryProgress*>& o);
void encode(bufferlist &bl) const;
void decode(bufferlist::const_iterator &bl);
ostream &print(ostream &out) const;
void dump(Formatter *f) const;
};
WRITE_CLASS_ENCODER(ObjectRecoveryProgress)
ostream& operator<<(ostream& out, const ObjectRecoveryProgress &prog);
struct PushReplyOp {
hobject_t soid;
static void generate_test_instances(list<PushReplyOp*>& o);
void encode(bufferlist &bl) const;
void decode(bufferlist::const_iterator &bl);
ostream &print(ostream &out) const;
void dump(Formatter *f) const;
uint64_t cost(CephContext *cct) const;
};
WRITE_CLASS_ENCODER(PushReplyOp)
ostream& operator<<(ostream& out, const PushReplyOp &op);
struct PullOp {
hobject_t soid;
ObjectRecoveryInfo recovery_info;
ObjectRecoveryProgress recovery_progress;
static void generate_test_instances(list<PullOp*>& o);
void encode(bufferlist &bl, uint64_t features) const;
void decode(bufferlist::const_iterator &bl);
ostream &print(ostream &out) const;
void dump(Formatter *f) const;
uint64_t cost(CephContext *cct) const;
};
WRITE_CLASS_ENCODER_FEATURES(PullOp)
ostream& operator<<(ostream& out, const PullOp &op);
struct PushOp {
hobject_t soid;
eversion_t version;
bufferlist data;
interval_set<uint64_t> data_included;
bufferlist omap_header;
map<string, bufferlist> omap_entries;
map<string, bufferlist> attrset;
ObjectRecoveryInfo recovery_info;
ObjectRecoveryProgress before_progress;
ObjectRecoveryProgress after_progress;
static void generate_test_instances(list<PushOp*>& o);
void encode(bufferlist &bl, uint64_t features) const;
void decode(bufferlist::const_iterator &bl);
ostream &print(ostream &out) const;
void dump(Formatter *f) const;
uint64_t cost(CephContext *cct) const;
};
WRITE_CLASS_ENCODER_FEATURES(PushOp)
ostream& operator<<(ostream& out, const PushOp &op);
/*
* summarize pg contents for purposes of a scrub
*/
struct ScrubMap {
struct object {
map<string,bufferptr> attrs;
uint64_t size;
__u32 omap_digest; ///< omap crc32c
__u32 digest; ///< data crc32c
bool negative:1;
bool digest_present:1;
bool omap_digest_present:1;
bool read_error:1;
bool stat_error:1;
bool ec_hash_mismatch:1;
bool ec_size_mismatch:1;
bool large_omap_object_found:1;
uint64_t large_omap_object_key_count = 0;
uint64_t large_omap_object_value_size = 0;
uint64_t object_omap_bytes = 0;
uint64_t object_omap_keys = 0;
object() :
// Init invalid size so it won't match if we get a stat EIO error
size(-1), omap_digest(0), digest(0),
negative(false), digest_present(false), omap_digest_present(false),
read_error(false), stat_error(false), ec_hash_mismatch(false),
ec_size_mismatch(false), large_omap_object_found(false) {}
void encode(bufferlist& bl) const;
void decode(bufferlist::const_iterator& bl);
void dump(Formatter *f) const;
static void generate_test_instances(list<object*>& o);
};
WRITE_CLASS_ENCODER(object)
map<hobject_t,object> objects;
eversion_t valid_through;
eversion_t incr_since;
bool has_large_omap_object_errors:1;
bool has_omap_keys:1;
void merge_incr(const ScrubMap &l);
void clear_from(const hobject_t& start) {
objects.erase(objects.lower_bound(start), objects.end());
}
void insert(const ScrubMap &r) {
objects.insert(r.objects.begin(), r.objects.end());
}
void swap(ScrubMap &r) {
using std::swap;
swap(objects, r.objects);
swap(valid_through, r.valid_through);
swap(incr_since, r.incr_since);
}
void encode(bufferlist& bl) const;
void decode(bufferlist::const_iterator& bl, int64_t pool=-1);
void dump(Formatter *f) const;
static void generate_test_instances(list<ScrubMap*>& o);
};
WRITE_CLASS_ENCODER(ScrubMap::object)
WRITE_CLASS_ENCODER(ScrubMap)
struct ScrubMapBuilder {
bool deep = false;
vector<hobject_t> ls;
size_t pos = 0;
int64_t data_pos = 0;
string omap_pos;
int ret = 0;
bufferhash data_hash, omap_hash; ///< accumulatinng hash value
uint64_t omap_keys = 0;
uint64_t omap_bytes = 0;
bool empty() {
return ls.empty();
}
bool done() {
return pos >= ls.size();
}
void reset() {
*this = ScrubMapBuilder();
}
bool data_done() {
return data_pos < 0;
}
void next_object() {
++pos;
data_pos = 0;
omap_pos.clear();
omap_keys = 0;
omap_bytes = 0;
}
friend ostream& operator<<(ostream& out, const ScrubMapBuilder& pos) {
out << "(" << pos.pos << "/" << pos.ls.size();
if (pos.pos < pos.ls.size()) {
out << " " << pos.ls[pos.pos];
}
if (pos.data_pos < 0) {
out << " byte " << pos.data_pos;
}
if (!pos.omap_pos.empty()) {
out << " key " << pos.omap_pos;
}
if (pos.deep) {
out << " deep";
}
if (pos.ret) {
out << " ret " << pos.ret;
}
return out << ")";
}
};
struct OSDOp {
ceph_osd_op op;
sobject_t soid;
bufferlist indata, outdata;
errorcode32_t rval;
OSDOp() : rval(0) {
// FIPS zeroization audit 20191115: this memset clean for security
memset(&op, 0, sizeof(ceph_osd_op));
}
/**
* split a bufferlist into constituent indata members of a vector of OSDOps
*
* @param ops [out] vector of OSDOps
* @param in [in] combined data buffer
*/
static void split_osd_op_vector_in_data(vector<OSDOp>& ops, bufferlist& in);
/**
* merge indata members of a vector of OSDOp into a single bufferlist
*
* Notably this also encodes certain other OSDOp data into the data
* buffer, including the sobject_t soid.
*
* @param ops [in] vector of OSDOps
* @param out [out] combined data buffer
*/
static void merge_osd_op_vector_in_data(vector<OSDOp>& ops, bufferlist& out);
/**
* split a bufferlist into constituent outdata members of a vector of OSDOps
*
* @param ops [out] vector of OSDOps
* @param in [in] combined data buffer
*/
static void split_osd_op_vector_out_data(vector<OSDOp>& ops, bufferlist& in);
/**
* merge outdata members of a vector of OSDOps into a single bufferlist
*
* @param ops [in] vector of OSDOps
* @param out [out] combined data buffer
*/
static void merge_osd_op_vector_out_data(vector<OSDOp>& ops, bufferlist& out);
/**
* Clear data as much as possible, leave minimal data for historical op dump
*
* @param ops [in] vector of OSDOps
*/
static void clear_data(vector<OSDOp>& ops);
};
ostream& operator<<(ostream& out, const OSDOp& op);
struct watch_item_t {
entity_name_t name;
uint64_t cookie;
uint32_t timeout_seconds;
entity_addr_t addr;
watch_item_t() : cookie(0), timeout_seconds(0) { }
watch_item_t(entity_name_t name, uint64_t cookie, uint32_t timeout,
const entity_addr_t& addr)
: name(name), cookie(cookie), timeout_seconds(timeout),
addr(addr) { }
void encode(bufferlist &bl, uint64_t features) const {
ENCODE_START(2, 1, bl);
encode(name, bl);
encode(cookie, bl);
encode(timeout_seconds, bl);
encode(addr, bl, features);
ENCODE_FINISH(bl);
}
void decode(bufferlist::const_iterator &bl) {
DECODE_START(2, bl);
decode(name, bl);
decode(cookie, bl);
decode(timeout_seconds, bl);
if (struct_v >= 2) {
decode(addr, bl);
}
DECODE_FINISH(bl);
}
};
WRITE_CLASS_ENCODER_FEATURES(watch_item_t)
struct obj_watch_item_t {
hobject_t obj;
watch_item_t wi;
};
/**
* obj list watch response format
*
*/
struct obj_list_watch_response_t {
list<watch_item_t> entries;
void encode(bufferlist& bl, uint64_t features) const {
ENCODE_START(1, 1, bl);
encode(entries, bl, features);
ENCODE_FINISH(bl);
}
void decode(bufferlist::const_iterator& bl) {
DECODE_START(1, bl);
decode(entries, bl);
DECODE_FINISH(bl);
}
void dump(Formatter *f) const {
f->open_array_section("entries");
for (list<watch_item_t>::const_iterator p = entries.begin(); p != entries.end(); ++p) {
f->open_object_section("watch");
f->dump_stream("watcher") << p->name;
f->dump_int("cookie", p->cookie);
f->dump_int("timeout", p->timeout_seconds);
f->open_object_section("addr");
p->addr.dump(f);
f->close_section();
f->close_section();
}
f->close_section();
}
static void generate_test_instances(list<obj_list_watch_response_t*>& o) {
entity_addr_t ea;
o.push_back(new obj_list_watch_response_t);
o.push_back(new obj_list_watch_response_t);
ea.set_type(entity_addr_t::TYPE_LEGACY);
ea.set_nonce(1000);
ea.set_family(AF_INET);
ea.set_in4_quad(0, 127);
ea.set_in4_quad(1, 0);
ea.set_in4_quad(2, 0);
ea.set_in4_quad(3, 1);
ea.set_port(1024);
o.back()->entries.push_back(watch_item_t(entity_name_t(entity_name_t::TYPE_CLIENT, 1), 10, 30, ea));
ea.set_nonce(1001);
ea.set_in4_quad(3, 2);
ea.set_port(1025);
o.back()->entries.push_back(watch_item_t(entity_name_t(entity_name_t::TYPE_CLIENT, 2), 20, 60, ea));
}
};
WRITE_CLASS_ENCODER_FEATURES(obj_list_watch_response_t)
struct clone_info {
snapid_t cloneid;
vector<snapid_t> snaps; // ascending
vector< pair<uint64_t,uint64_t> > overlap;
uint64_t size;
clone_info() : cloneid(CEPH_NOSNAP), size(0) {}
void encode(bufferlist& bl) const {
ENCODE_START(1, 1, bl);
encode(cloneid, bl);
encode(snaps, bl);
encode(overlap, bl);
encode(size, bl);
ENCODE_FINISH(bl);
}
void decode(bufferlist::const_iterator& bl) {
DECODE_START(1, bl);
decode(cloneid, bl);
decode(snaps, bl);
decode(overlap, bl);
decode(size, bl);
DECODE_FINISH(bl);
}
void dump(Formatter *f) const {
if (cloneid == CEPH_NOSNAP)
f->dump_string("cloneid", "HEAD");
else
f->dump_unsigned("cloneid", cloneid.val);
f->open_array_section("snapshots");
for (vector<snapid_t>::const_iterator p = snaps.begin(); p != snaps.end(); ++p) {
f->open_object_section("snap");
f->dump_unsigned("id", p->val);
f->close_section();
}
f->close_section();
f->open_array_section("overlaps");
for (vector< pair<uint64_t,uint64_t> >::const_iterator q = overlap.begin();
q != overlap.end(); ++q) {
f->open_object_section("overlap");
f->dump_unsigned("offset", q->first);
f->dump_unsigned("length", q->second);
f->close_section();
}
f->close_section();
f->dump_unsigned("size", size);
}
static void generate_test_instances(list<clone_info*>& o) {
o.push_back(new clone_info);
o.push_back(new clone_info);
o.back()->cloneid = 1;
o.back()->snaps.push_back(1);
o.back()->overlap.push_back(pair<uint64_t,uint64_t>(0,4096));
o.back()->overlap.push_back(pair<uint64_t,uint64_t>(8192,4096));
o.back()->size = 16384;
o.push_back(new clone_info);
o.back()->cloneid = CEPH_NOSNAP;
o.back()->size = 32768;
}
};
WRITE_CLASS_ENCODER(clone_info)
/**
* obj list snaps response format
*
*/
struct obj_list_snap_response_t {
vector<clone_info> clones; // ascending
snapid_t seq;
void encode(bufferlist& bl) const {
ENCODE_START(2, 1, bl);
encode(clones, bl);
encode(seq, bl);
ENCODE_FINISH(bl);
}
void decode(bufferlist::const_iterator& bl) {
DECODE_START(2, bl);
decode(clones, bl);
if (struct_v >= 2)
decode(seq, bl);
else
seq = CEPH_NOSNAP;
DECODE_FINISH(bl);
}
void dump(Formatter *f) const {
f->open_array_section("clones");
for (vector<clone_info>::const_iterator p = clones.begin(); p != clones.end(); ++p) {
f->open_object_section("clone");
p->dump(f);
f->close_section();
}
f->dump_unsigned("seq", seq);
f->close_section();
}
static void generate_test_instances(list<obj_list_snap_response_t*>& o) {
o.push_back(new obj_list_snap_response_t);
o.push_back(new obj_list_snap_response_t);
clone_info cl;
cl.cloneid = 1;
cl.snaps.push_back(1);
cl.overlap.push_back(pair<uint64_t,uint64_t>(0,4096));
cl.overlap.push_back(pair<uint64_t,uint64_t>(8192,4096));
cl.size = 16384;
o.back()->clones.push_back(cl);
cl.cloneid = CEPH_NOSNAP;
cl.snaps.clear();
cl.overlap.clear();
cl.size = 32768;
o.back()->clones.push_back(cl);
o.back()->seq = 123;
}
};
WRITE_CLASS_ENCODER(obj_list_snap_response_t)
// PromoteCounter
struct PromoteCounter {
std::atomic<unsigned long long> attempts{0};
std::atomic<unsigned long long> objects{0};
std::atomic<unsigned long long> bytes{0};
void attempt() {
attempts++;
}
void finish(uint64_t size) {
objects++;
bytes += size;
}
void sample_and_attenuate(uint64_t *a, uint64_t *o, uint64_t *b) {
*a = attempts;
*o = objects;
*b = bytes;
attempts = *a / 2;
objects = *o / 2;
bytes = *b / 2;
}
};
struct pool_pg_num_history_t {
/// last epoch updated
epoch_t epoch = 0;
/// poolid -> epoch -> pg_num
map<int64_t,map<epoch_t,uint32_t>> pg_nums;
/// pair(epoch, poolid)
set<pair<epoch_t,int64_t>> deleted_pools;
void log_pg_num_change(epoch_t epoch, int64_t pool, uint32_t pg_num) {
pg_nums[pool][epoch] = pg_num;
}
void log_pool_delete(epoch_t epoch, int64_t pool) {
deleted_pools.insert(make_pair(epoch, pool));
}
/// prune history based on oldest osdmap epoch in the cluster
void prune(epoch_t oldest_epoch) {
auto i = deleted_pools.begin();
while (i != deleted_pools.end()) {
if (i->first >= oldest_epoch) {
break;
}
pg_nums.erase(i->second);
i = deleted_pools.erase(i);
}
for (auto& j : pg_nums) {
auto k = j.second.lower_bound(oldest_epoch);
// keep this and the entry before it (just to be paranoid)
if (k != j.second.begin()) {
--k;
j.second.erase(j.second.begin(), k);
}
}
}
void encode(bufferlist& bl) const {
ENCODE_START(1, 1, bl);
encode(epoch, bl);
encode(pg_nums, bl);
encode(deleted_pools, bl);
ENCODE_FINISH(bl);
}
void decode(bufferlist::const_iterator& p) {
DECODE_START(1, p);
decode(epoch, p);
decode(pg_nums, p);
decode(deleted_pools, p);
DECODE_FINISH(p);
}
void dump(Formatter *f) const {
f->dump_unsigned("epoch", epoch);
f->open_object_section("pools");
for (auto& i : pg_nums) {
f->open_object_section("pool");
f->dump_unsigned("pool_id", i.first);
f->open_array_section("changes");
for (auto& j : i.second) {
f->open_object_section("change");
f->dump_unsigned("epoch", j.first);
f->dump_unsigned("pg_num", j.second);
f->close_section();
}
f->close_section();
f->close_section();
}
f->close_section();
f->open_array_section("deleted_pools");
for (auto& i : deleted_pools) {
f->open_object_section("deletion");
f->dump_unsigned("pool_id", i.second);
f->dump_unsigned("epoch", i.first);
f->close_section();
}
f->close_section();
}
static void generate_test_instances(list<pool_pg_num_history_t*>& ls) {
ls.push_back(new pool_pg_num_history_t);
}
friend ostream& operator<<(ostream& out, const pool_pg_num_history_t& h) {
return out << "pg_num_history(e" << h.epoch
<< " pg_nums " << h.pg_nums
<< " deleted_pools " << h.deleted_pools
<< ")";
}
};
WRITE_CLASS_ENCODER(pool_pg_num_history_t)
// omap specific stats
struct omap_stat_t {
int large_omap_objects;
int64_t omap_bytes;
int64_t omap_keys;
};
#endif
|