9 files changed, 193 insertions, 0 deletions
diff --git a/qa/suites/rados/mgr/tasks/.qa b/qa/suites/rados/mgr/tasks/.qa
new file mode 120000
index 000000000..a602a0353
--- /dev/null
+++ b/qa/suites/rados/mgr/tasks/.qa
@@ -0,0 +1 @@
+../.qa/
+\ No newline at end of file
diff --git a/qa/suites/rados/mgr/tasks/crash.yaml b/qa/suites/rados/mgr/tasks/crash.yaml
new file mode 100644
index 000000000..9d2ba535e
--- /dev/null
+++ b/qa/suites/rados/mgr/tasks/crash.yaml
@@ -0,0 +1,18 @@
+
+tasks:
+  - install:
+  - ceph:
+      # tests may leave mgrs broken, so don't try and call into them
+      # to invoke e.g. pg dump during teardown.
+      wait-for-scrub: false
+      log-ignorelist:
+        - overall HEALTH_
+        - \(MGR_DOWN\)
+        - \(PG_
+        - \(RECENT_CRASH\)
+        - replacing it with standby
+        - No standby daemons available
+        - \(POOL_APP_NOT_ENABLED\)
+  - cephfs_test_runner:
+      modules:
+        - tasks.mgr.test_crash
diff --git a/qa/suites/rados/mgr/tasks/failover.yaml b/qa/suites/rados/mgr/tasks/failover.yaml
new file mode 100644
index 000000000..6d1e0d557
--- /dev/null
+++ b/qa/suites/rados/mgr/tasks/failover.yaml
@@ -0,0 +1,17 @@
+
+tasks:
+  - install:
+  - ceph:
+      # tests may leave mgrs broken, so don't try and call into them
+      # to invoke e.g. pg dump during teardown.
+      wait-for-scrub: false
+      log-ignorelist:
+        - overall HEALTH_
+        - \(MGR_DOWN\)
+        - \(PG_
+        - replacing it with standby
+        - No standby daemons available
+        - \(POOL_APP_NOT_ENABLED\)
+  - cephfs_test_runner:
+      modules:
+        - tasks.mgr.test_failover
diff --git a/qa/suites/rados/mgr/tasks/insights.yaml b/qa/suites/rados/mgr/tasks/insights.yaml
new file mode 100644
index 000000000..f7c82cf7f
--- /dev/null
+++ b/qa/suites/rados/mgr/tasks/insights.yaml
@@ -0,0 +1,20 @@
+
+tasks:
+  - install:
+  - ceph:
+      # tests may leave mgrs broken, so don't try and call into them
+      # to invoke e.g. pg dump during teardown.
+      wait-for-scrub: false
+      log-ignorelist:
+        - overall HEALTH_
+        - \(MGR_DOWN\)
+        - \(MGR_INSIGHTS_WARNING\)
+        - \(insights_health_check
+        - \(PG_
+        - \(RECENT_CRASH\)
+        - replacing it with standby
+        - No standby daemons available
+        - \(POOL_APP_NOT_ENABLED\)
+  - cephfs_test_runner:
+      modules:
+        - tasks.mgr.test_insights
diff --git a/qa/suites/rados/mgr/tasks/module_selftest.yaml b/qa/suites/rados/mgr/tasks/module_selftest.yaml
new file mode 100644
index 000000000..4403d9fff
--- /dev/null
+++ b/qa/suites/rados/mgr/tasks/module_selftest.yaml
@@ -0,0 +1,28 @@
+
+tasks:
+  - install:
+  - ceph:
+      # tests may leave mgrs broken, so don't try and call into them
+      # to invoke e.g. pg dump during teardown.
+      wait-for-scrub: false
+      log-ignorelist:
+        - overall HEALTH_
+        - \(MGR_DOWN\)
+        - \(PG_
+        - replacing it with standby
+        - No standby daemons available
+        - Reduced data availability
+        - Degraded data redundancy
+        - objects misplaced
+        - Synthetic exception in serve
+        - influxdb python module not found
+        - \(MGR_ZABBIX_
+        - foo bar
+        - Failed to open Telegraf
+        - evicting unresponsive client
+        - 1 mgr modules have recently crashed \(RECENT_MGR_MODULE_CRASH\)
+        - \(POOL_APP_NOT_ENABLED\)
+  - cephfs_test_runner:
+      modules:
+        - tasks.mgr.test_module_selftest
+      fail_on_skip: false
diff --git a/qa/suites/rados/mgr/tasks/per_module_finisher_stats.yaml b/qa/suites/rados/mgr/tasks/per_module_finisher_stats.yaml
new file mode 100644
index 000000000..de1d592df
--- /dev/null
+++ b/qa/suites/rados/mgr/tasks/per_module_finisher_stats.yaml
@@ -0,0 +1,45 @@
+tasks:
+  - install:
+  - ceph:
+      wait-for-scrub: false
+      log-ignorelist:
+        - \(POOL_APP_NOT_ENABLED\)
+  - check-counter:
+      counters:
+        mgr:
+            - name: "finisher-balancer.complete_latency.avgcount"
+              min: 1
+            - name: "finisher-balancer.queue_len"
+              expected_val: 0
+            - name: "finisher-crash.complete_latency.avgcount"
+              min: 2
+            - name: "finisher-crash.queue_len"
+              expected_val: 0
+            - name: "finisher-devicehealth.complete_latency.avgcount"
+              min: 1
+            - name: "finisher-devicehealth.queue_len"
+              expected_val: 0
+            - name: "finisher-iostat.complete_latency.avgcount"
+              min: 1
+            - name: "finisher-iostat.queue_len"
+              expected_val: 0
+            - name: "finisher-pg_autoscaler.complete_latency.avgcount"
+              min: 1
+            - name: "finisher-pg_autoscaler.queue_len"
+              expected_val: 0
+            - name: "finisher-progress.complete_latency.avgcount"
+              min: 2
+            - name: "finisher-progress.queue_len"
+              expected_val: 0
+            - name: "finisher-status.complete_latency.avgcount"
+              min: 2
+            - name: "finisher-status.queue_len"
+              expected_val: 0
+            - name: "finisher-telemetry.complete_latency.avgcount"
+              min: 2
+            - name: "finisher-telemetry.queue_len"
+              expected_val: 0
+  - workunit:
+      clients:
+        client.0:
+          - mgr/test_per_module_finisher.sh
diff --git a/qa/suites/rados/mgr/tasks/progress.yaml b/qa/suites/rados/mgr/tasks/progress.yaml
new file mode 100644
index 000000000..183a9a29a
--- /dev/null
+++ b/qa/suites/rados/mgr/tasks/progress.yaml
@@ -0,0 +1,30 @@
+overrides:
+  ceph:
+    conf:
+      osd:
+        osd mclock profile: high_recovery_ops
+tasks:
+  - install:
+  - ceph:
+      config:
+        global:
+          osd pool default size : 3
+          osd pool default min size : 2
+      # tests may leave mgrs broken, so don't try and call into them
+      # to invoke e.g. pg dump during teardown.
+      wait-for-scrub: false
+      log-ignorelist:
+        - overall HEALTH_
+        - \(MGR_DOWN\)
+        - \(MDS_ALL_DOWN\)
+        - \(MDS_UP_LESS_THAN_MAX\)
+        - \(FS_WITH_FAILED_MDS\)
+        - \(FS_DEGRADED\)
+        - \(PG_
+        - \(OSDMAP_FLAGS\)
+        - replacing it with standby
+        - No standby daemons available
+        - \(POOL_APP_NOT_ENABLED\)
+  - cephfs_test_runner:
+      modules:
+        - tasks.mgr.test_progress
diff --git a/qa/suites/rados/mgr/tasks/prometheus.yaml b/qa/suites/rados/mgr/tasks/prometheus.yaml
new file mode 100644
index 000000000..fd0e23a35
--- /dev/null
+++ b/qa/suites/rados/mgr/tasks/prometheus.yaml
@@ -0,0 +1,17 @@
+
+tasks:
+  - install:
+  - ceph:
+      # tests may leave mgrs broken, so don't try and call into them
+      # to invoke e.g. pg dump during teardown.
+      wait-for-scrub: false
+      log-ignorelist:
+        - overall HEALTH_
+        - \(MGR_DOWN\)
+        - \(PG_
+        - replacing it with standby
+        - No standby daemons available
+        - \(POOL_APP_NOT_ENABLED\)
+  - cephfs_test_runner:
+      modules:
+        - tasks.mgr.test_prometheus
diff --git a/qa/suites/rados/mgr/tasks/workunits.yaml b/qa/suites/rados/mgr/tasks/workunits.yaml
new file mode 100644
index 000000000..a48274033
--- /dev/null
+++ b/qa/suites/rados/mgr/tasks/workunits.yaml
@@ -0,0 +1,17 @@
+tasks:
+  - install:
+  - ceph:
+      # tests may leave mgrs broken, so don't try and call into them
+      # to invoke e.g. pg dump during teardown.
+      wait-for-scrub: false
+      log-ignorelist:
+        - overall HEALTH_
+        - \(MGR_DOWN\)
+        - \(PG_
+        - replacing it with standby
+        - No standby daemons available
+        - \(POOL_APP_NOT_ENABLED\)
+  - workunit:
+      clients:
+        client.0:
+          - mgr/test_localpool.sh