[
https://issues.apache.org/jira/browse/CASSANDRA-10754?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15022199#comment-15022199
]
Sylvain Lebresne commented on CASSANDRA-10754:
----------------------------------------------
This is a problem with the test: the invariant is good to keep but the test
clearly don't respect it since it re-generate a timestamp for each query. We
can easily generate the "nowInSeconds" value once during static initialization
of the class and use that for all commands as far as this test is concerned.
[~jkni]: care to provide a patch for that?
> Failure in QueryPagerTest.multiQueryTest due to assertion in
> SinglePartitionReadCommand$Group
> ---------------------------------------------------------------------------------------------
>
> Key: CASSANDRA-10754
> URL: https://issues.apache.org/jira/browse/CASSANDRA-10754
> Project: Cassandra
> Issue Type: Bug
> Reporter: Joel Knighton
>
> Test failure
> [here|http://cassci.datastax.com/view/Dev/view/jkni/job/jkni-9510-3.0-testall/lastCompletedBuild/testReport/org.apache.cassandra.service/QueryPagerTest/multiQueryTest/].
> It's on a 3.0 branch for a issue, but the patch only affects the
> implementation of the {{assassinate}} command and wouldn't affect this unit
> test.
> The assertion failure is in the construction of a Group of
> SinglePartitionReadCommands, where the value of {{nowInSec}} for each command
> does not match.
> I'm not sure if 1) this invariant isn't valuable, 2) this invariant is
> valuable but the unit test does not accurately reflect how a group would be
> created, or 3) this is a genuine bug.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
