[
https://issues.apache.org/jira/browse/CASSANDRA-11289?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Sam Tunnicliffe updated CASSANDRA-11289:
----------------------------------------
Status: Patch Available (was: Open)
I was able to repro this approximately 1 time out of every 40 runs and it seems
that the problem is simply that the trace events that the test checks for
haven't been populated. Adding a small backoff & retry to the check solves that
problem, at least locally, though it's obviously not foolproof as the lag
writing the trace events may vary in the CI environment. Despite this being an
imperfect solution, I don't have any much better suggestion so I've opened a
[dtest PR|https://github.com/riptano/cassandra-dtest/pull/846] to add it & I'll
close this when that's merged.
> dtest failure in
> secondary_indexes_test.TestSecondaryIndexes.test_only_coordinator_chooses_index_for_query
> ----------------------------------------------------------------------------------------------------------
>
> Key: CASSANDRA-11289
> URL: https://issues.apache.org/jira/browse/CASSANDRA-11289
> Project: Cassandra
> Issue Type: Test
> Reporter: Russ Hatch
> Assignee: Sam Tunnicliffe
> Labels: dtest
>
> example failure:
> http://cassci.datastax.com/job/trunk_novnode_dtest/297/testReport/secondary_indexes_test/TestSecondaryIndexes/test_only_coordinator_chooses_index_for_query
> Failed on CassCI build trunk_novnode_dtest #297
> Failing intermittently, error looks like this:
> {noformat}
> Expected to find between 1 and 200 trace events matching Executing read on
> ks.cf using index b_index from 127.0.0.1, but actually found 0. (Full counts:
> {'127.0.0.3': 1, '127.0.0.2': 1, '127.0.0.1': 0})
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
