Closed zimnx closed 2 months ago
This was on -clusterip job which has slower persistent network attached ssds. We might want to reevaluate whether we still want to use them.
-clusterip
https://prow.scylla-operator.scylladb.com/view/gs/scylla-operator-prow/pr-logs/pull/scylladb_scylla-operator/1991/pull-scylla-operator-master-e2e-gke-parallel-clusterip/1806721597633990656
[FAILED] Unexpected error: <*fmt.wrapError | 0xc0006967a0>: can't insert data: Operation timed out for 58qkpsqt.test - received only 1 responses from 2 CL=ALL. { msg: "can't insert data: Operation timed out for 58qkpsqt.test - received only 1 responses from 2 CL=ALL.", err: <*gocql.RequestErrWriteTimeout | 0xc0004f4380>{ errorFrame: { frameHeader: {version: 132, flags: 0, stream: 576, op: 0, length: 104, warnings: nil}, code: 4352, message: "Operation timed out for 58qkpsqt.test - received only 1 responses from 2 CL=ALL.", }, Consistency: 5, Received: 1, BlockFor: 2, WriteType: "SIMPLE", }, } occurred In [It] at: github.com/scylladb/scylla-operator/test/e2e/set/scyllacluster/verify.go:312 @ 06/28/24 16:30:32.464
https://prow.scylla-operator.scylladb.com/view/gs/scylla-operator-prow/pr-logs/pull/scylladb_scylla-operator/1991/pull-scylla-operator-master-e2e-gke-parallel-clusterip/1806749115925663744
We don' have a choice on some platforms - preferably we'd adjust the timeout or concurrency based o what exactly goes so slow. /priority important-longterm /triage accepted
https://prow.scylla-operator.scylladb.com/view/gs/scylla-operator-prow/pr-logs/pull/scylladb_scylla-operator/1871/pull-scylla-operator-master-e2e-gke-parallel-clusterip/1808031823570145280#1:test-build-log.txt%3A1205
https://prow.scylla-operator.scylladb.com/view/gs/scylla-operator-prow/pr-logs/pull/scylladb_scylla-operator/1971/pull-scylla-operator-master-e2e-gke-parallel-clusterip/1810558454176157696#1:test-build-log.txt%3A962
https://prow.scylla-operator.scylladb.com/view/gs/scylla-operator-prow/pr-logs/pull/scylladb_scylla-operator/1971/pull-scylla-operator-master-e2e-gke-parallel-clusterip/1810586982485594112#1:test-build-log.txt%3A750
https://prow.scylla-operator.scylladb.com/view/gs/scylla-operator-prow/pr-logs/pull/scylladb_scylla-operator/1971/pull-scylla-operator-master-e2e-gke-parallel-clusterip/1810605342397042688#1:test-build-log.txt%3A1258
This was on
-clusterip
job which has slower persistent network attached ssds. We might want to reevaluate whether we still want to use them.Link to the job that flaked.
https://prow.scylla-operator.scylladb.com/view/gs/scylla-operator-prow/pr-logs/pull/scylladb_scylla-operator/1991/pull-scylla-operator-master-e2e-gke-parallel-clusterip/1806721597633990656
Snippet of what failed.