Merging upstream version 18.2.3.

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

6 files changed, 59 insertions, 21 deletions

diff --git a/src/common/options/cephfs-mirror.yaml.in b/src/common/options/cephfs-mirror.yaml.in
index 78f86dfb1..f82616187 100644
--- a/src/common/options/cephfs-mirror.yaml.in
+++ b/src/common/options/cephfs-mirror.yaml.in
@@ -91,4 +91,15 @@ options:
   default: 10
   services:
   - cephfs-mirror
-  min: 0
\ No newline at end of file
+  min: 0
+- name: cephfs_mirror_perf_stats_prio
+  type: int
+  level: advanced
+  desc: Priority level for mirror daemon replication perf counters
+  long_desc: The daemon will send perf counter data to the manager daemon if the priority
+    is not lower than mgr_stats_threshold.
+  default: 5
+  services:
+  - cephfs-mirror
+  min: 0
+  max: 11
diff --git a/src/common/options/global.yaml.in b/src/common/options/global.yaml.in
index fa426a115..f0eaedf5a 100644
--- a/src/common/options/global.yaml.in
+++ b/src/common/options/global.yaml.in
@@ -4964,7 +4964,7 @@ options:
   type: str
   level: advanced
   desc: Full set of rocksdb settings to override
-  default: compression=kNoCompression,max_write_buffer_number=64,min_write_buffer_number_to_merge=6,compaction_style=kCompactionStyleLevel,write_buffer_size=16777216,max_background_jobs=4,level0_file_num_compaction_trigger=8,max_bytes_for_level_base=1073741824,max_bytes_for_level_multiplier=8,compaction_readahead_size=2MB,max_total_wal_size=1073741824,writable_file_max_buffer_size=0
+  default: compression=kLZ4Compression,max_write_buffer_number=64,min_write_buffer_number_to_merge=6,compaction_style=kCompactionStyleLevel,write_buffer_size=16777216,max_background_jobs=4,level0_file_num_compaction_trigger=8,max_bytes_for_level_base=1073741824,max_bytes_for_level_multiplier=8,compaction_readahead_size=2MB,max_total_wal_size=1073741824,writable_file_max_buffer_size=0
   with_legacy: true
 - name: bluestore_rocksdb_options_annex
   type: str
diff --git a/src/common/options/mds-client.yaml.in b/src/common/options/mds-client.yaml.in
index 4e599d4cf..1f7600dee 100644
--- a/src/common/options/mds-client.yaml.in
+++ b/src/common/options/mds-client.yaml.in
@@ -315,6 +315,8 @@ options:
   default: true
   services:
   - mds_client
+  flags:
+  - startup
   with_legacy: true
 - name: client_force_lazyio
   type: bool
diff --git a/src/common/options/mds.yaml.in b/src/common/options/mds.yaml.in
index 6eb0702fc..6234b96cd 100644
--- a/src/common/options/mds.yaml.in
+++ b/src/common/options/mds.yaml.in
@@ -65,15 +65,6 @@ options:
   - mds
   flags:
   - runtime
-# max xattr kv pairs size for each dir/file
-- name: mds_max_xattr_pairs_size
-  type: size
-  level: advanced
-  desc: maximum aggregate size of extended attributes on a file
-  default: 64_K
-  services:
-  - mds
-  with_legacy: true
 - name: mds_cache_trim_interval
   type: secs
   level: advanced
@@ -1175,6 +1166,24 @@ options:
   services:
   - mds
   with_legacy: true
+# Max number of slow ops to track
+- name: mds_op_history_slow_op_size
+  type: uint
+  level: advanced
+  desc: maximum size for list of historical slow operations
+  default: 20
+  services:
+  - mds
+  with_legacy: true
+# Track the op if over this threshold
+- name: mds_op_history_slow_op_threshold
+  type: uint
+  level: advanced
+  desc: track the op if over this threshold
+  default: 10
+  services:
+  - mds
+  with_legacy: true
 # how many seconds old makes an op complaint-worthy
 - name: mds_op_complaint_time
   type: float
@@ -1533,4 +1542,16 @@ options:
   services:
   - mds
   flags:
+  - runtime  
+- name: defer_client_eviction_on_laggy_osds
+  type: bool
+  level: advanced
+  desc: Do not evict client if any osd is laggy
+  long_desc: Laggy OSD(s) can make clients laggy or unresponsive, this can
+    lead to their eviction, this option once enabled can help defer client
+    eviction.
+  default: false
+  services:
+  - mds
+  flags:
   - runtime
diff --git a/src/common/options/osd.yaml.in b/src/common/options/osd.yaml.in
index 7291ce11d..a9596cec1 100644
--- a/src/common/options/osd.yaml.in
+++ b/src/common/options/osd.yaml.in
@@ -182,7 +182,7 @@ options:
   desc: Maximum concurrent scrubs on a single OSD
   fmt_desc: The maximum number of simultaneous scrub operations for
     a Ceph OSD Daemon.
-  default: 1
+  default: 3
   with_legacy: true
 - name: osd_scrub_during_recovery
   type: bool
@@ -895,13 +895,13 @@ options:
   desc: Do not store full-object checksums if the backend (bluestore) does its own
     checksums.  Only usable with all BlueStore OSDs.
   default: false
-# PrioritzedQueue (prio), Weighted Priority Queue (wpq ; default),
-# mclock_opclass, mclock_client, or debug_random. "mclock_opclass"
-# and "mclock_client" are based on the mClock/dmClock algorithm
-# (Gulati, et al. 2010). "mclock_opclass" prioritizes based on the
-# class the operation belongs to. "mclock_client" does the same but
-# also works to ienforce fairness between clients. "debug_random"
-# chooses among all four with equal probability.
+# Weighted Priority Queue (wpq), mClock Scheduler (mclock_scheduler: default)
+# or debug_random. "mclock_scheduler" is based on the mClock/dmClock
+# algorithm (Gulati, et al. 2010). "mclock_scheduler" prioritizes based on
+# the class the operation belongs to. "wpq" dequeues ops based on their
+# priorities. "debug_random" chooses among the two with equal probability.
+# Note: PrioritzedQueue (prio) implementation is not used for scheduling ops
+# within OSDs and is therefore not listed.
 - name: osd_op_queue
   type: str
   level: advanced
diff --git a/src/common/options/rgw.yaml.in b/src/common/options/rgw.yaml.in
index 241632a22..9c7f91f9e 100644
--- a/src/common/options/rgw.yaml.in
+++ b/src/common/options/rgw.yaml.in
@@ -233,7 +233,7 @@ options:
   long_desc: The lifecycle maintenance thread is responsible for lifecycle related
     maintenance work. The thread itself can be disabled, but in order for lifecycle
     to work correctly, at least one RGW in each zone needs to have this thread running.
-    Havingthe thread enabled on multiple RGW processes within the same zone can spread
+    Having the thread enabled on multiple RGW processes within the same zone can spread
     some of the maintenance work between them.
   default: true
   services:
@@ -359,7 +359,11 @@ options:
   type: str
   level: advanced
   desc: Lifecycle allowed work time
-  long_desc: Local time window in which the lifecycle maintenance thread can work.
+  long_desc: Local time window in which the lifecycle maintenance thread can work. It expects
+    24-hour time notation. For example, "00:00-23:59" means starting at midnight lifecycle
+    is allowed to run for the whole day (24 hours). When lifecycle completes, it waits for the
+    next maintenance window. In this example, if it completes at 01:00, it will resume processing
+    23 hours later at the following midnight.
   default: 00:00-06:00
   services:
   - rgw