Details
-
Bug
-
Resolution: Fixed
-
Critical
-
Cheshire-Cat
-
Untriaged
-
-
1
-
Unknown
Description
Build : 7.0.0-3154
Test : -test tests/2i/test_idx_steady_state_cheshire_cat.yml -scope tests/2i/scope_idx_cheshire_cat_moi.yml
GSI Storage Mode : MOI
Scale : 2
Memcached got OOM killed on the KV node 172.23.96.18 due to the projector memory consumption issue MB-41422.
As a result of this, there was a KV rollback. The memcached logs on 172.23.96.18 has messages like these -
2020-09-29T19:40:56.334689-07:00 WARNING 92: (bucket3) DCP (Producer) eq_dcpq:secidx:proj-bucket3-MAINT_STREAM_TOPIC_aab3f073fd2d8bff0bf3ca584ad910ad-7555551877666133025/0 - (vb:256) Stream request requires rollback to seqno:29974 because consumer ahead of producer - producer upper at 29974. Client requested seqnos:{29975,18446744073709551615} snapshot:{29968,29975} uuid:255571248687441
|
On the indexer log on 172.23.120.77, the following msg can be seen -
2020-09-29T19:40:56.040-07:00 [Info] Timekeeper::sendRestartMsg Received KV Repair Msg For Stream MAINT_STREAM KeyspaceId bucket1. Attempting Stream Repair.
|
And then, few mins later, on 172.23.120.77, indexer crashes with the following error :
2020-09-29T19:45:22.450-07:00 [Error] Indexer::sendStreamUpdateForBuildIndex Unexpected Rollback from Projector during Initial Stream Request &{3 bucket2:scope_6:coll_-2 0xc0294a6dc0 0 0}
|
panic: KV Rollback Received For Initial Build Request
|
|
goroutine 135288670 [running]:
|
github.com/couchbase/indexing/secondary/common.CrashOnError(...)
|
/home/couchbase/jenkins/workspace/couchbase-server-unix/goproj/src/github.com/couchbase/indexing/secondary/common/util.go:413
|
github.com/couchbase/indexing/secondary/indexer.(*indexer).sendStreamUpdateForBuildIndex.func1(0xc000077180, 0x3, 0xc00578d960, 0x17, 0xc000188780, 0x5, 0x8, 0x4, 0x1411e80, 0xc01919e240, ...)
|
/home/couchbase/jenkins/workspace/couchbase-server-unix/goproj/src/github.com/couchbase/indexing/secondary/indexer/indexer.go:4168 +0xf1e
|
created by github.com/couchbase/indexing/secondary/indexer.(*indexer).sendStreamUpdateForBuildIndex
|
/home/couchbase/jenkins/workspace/couchbase-server-unix/goproj/src/github.com/couchbase/indexing/secondary/indexer/indexer.go:4116 +0xf63
|
Indexer also crashed on other nodes around the same time. Following excerpt is from eagle-eye :
2020-09-29 20:08:00,866 - systestmon - WARNING - *** 1 occurences of exited with status keyword found on 172.23.120.77 ***
|
2020-09-29 20:08:00,872 - systestmon - DEBUG - [user:info,2020-09-29T19:45:22.864-07:00,ns_1@172.23.120.77:<0.11559.0>:ns_log:crash_consumption_loop:69]Service 'indexer' exited with status 2. Restarting. Messages:
|
2020-09-29 20:08:02,215 - systestmon - WARNING - *** 1 occurences of exited with status keyword found on 172.23.121.77 ***
|
2020-09-29 20:08:02,225 - systestmon - DEBUG - [user:info,2020-09-29T19:45:08.852-07:00,ns_1@172.23.121.77:<0.15443.0>:ns_log:crash_consumption_loop:69]Service 'indexer' exited with status 2. Restarting. Messages:
|
2020-09-29 20:08:02,900 - systestmon - WARNING - *** 1 occurences of exited with status keyword found on 172.23.123.24 ***
|
2020-09-29 20:08:02,907 - systestmon - DEBUG - [user:info,2020-09-29T19:45:17.881-07:00,ns_1@172.23.123.24:<0.13282.0>:ns_log:crash_consumption_loop:69]Service 'indexer' exited with status 2. Restarting. Messages:
|
2020-09-29 20:08:03,577 - systestmon - WARNING - *** 1 occurences of exited with status keyword found on 172.23.123.25 ***
|
2020-09-29 20:08:03,581 - systestmon - DEBUG - [user:info,2020-09-29T19:45:15.660-07:00,ns_1@172.23.123.25:<0.12149.0>:ns_log:crash_consumption_loop:69]Service 'indexer' exited with status 2. Restarting. Messages:
|
2020-09-29 20:08:04,254 - systestmon - WARNING - *** 1 occurences of exited with status keyword found on 172.23.123.26 ***
|
2020-09-29 20:08:04,263 - systestmon - DEBUG - [user:info,2020-09-29T19:46:01.083-07:00,ns_1@172.23.123.26:<0.12405.0>:ns_log:crash_consumption_loop:69]Service 'indexer' exited with status 2. Restarting. Messages:
|