Details
-
Bug
-
Resolution: Fixed
-
Critical
-
6.5.0
-
Triaged
-
No
Description
Build 6.5.0-4926
Observed following panic while running performance test experiments suggested by Varun Velamuri to understand incremental build timings.
In this experiment we have 1B docs, create 4 indexes on fields of ~128 byte length and then do incremental builds. This panic observed while building first index. We used plasma storage and supposed to go ~20% RR.
Job- http://perf.jenkins.couchbase.com/job/secondary-plasma-dgm-dev/65
Indexer logs- https://s3-us-west-2.amazonaws.com/perf-artifacts/jenkins-secondary-plasma-dgm-dev-65/172.23.97.10.zip
Panic-
panic: runtime error: integer divide by zero
|
goroutine 5972 [running]:
|
panic(0xf8b860, 0xc420016090)
|
/home/couchbase/.cbdepscache/exploded/x86_64/go-1.7.6/go/src/runtime/panic.go:500 +0x1a1
|
github.com/couchbase/plasma.(*Plasma).adjustSMOPageLimits(0xc423a15900)
|
goproj/src/github.com/couchbase/plasma/plasma.go:850 +0x9e1
|
github.com/couchbase/plasma.(*Plasma).monitorStatistics(0xc423a15900)
|
goproj/src/github.com/couchbase/plasma/plasma.go:958 +0x1a5
|
created by github.com/couchbase/plasma.New
|
goproj/src/github.com/couchbase/plasma/plasma.go:696 +0x1308
|
Attachments
Issue Links
- backports to
-
MB-41643 BP 6.6.1 MB-37225 Panic in indexer logs while running GSI performance experiments
- Closed