[
https://issues.apache.org/jira/browse/CASSANDRA-10938?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15085739#comment-15085739
]
Sylvain Lebresne commented on CASSANDRA-10938:
----------------------------------------------
I agree with changing to CHM here if that fixes Windows and cassci as
performance is not terribly critical in that case (important but not critical).
It does very much look like we're hitting either a very bad case or a bug in
NBHM. From a very quick look at the source, the methods that takes all the CPU
have some CAS loops so if lots of threads are trying to access the same slot
(which sounds to be the case if I understand the failing test correctly) while
a resizing is going on, then I suppose crappy scheduling could lead us there.
That bad behavior is a bit worrisome in the sense that we use NBHM in other
places, but it's also the first time we see this behavior so it's unclear the
ROI of investigating too much is worth it right now.
> test_bulk_round_trip_blogposts is failing occasionally
> ------------------------------------------------------
>
> Key: CASSANDRA-10938
> URL: https://issues.apache.org/jira/browse/CASSANDRA-10938
> Project: Cassandra
> Issue Type: Sub-task
> Components: Tools
> Reporter: Stefania
> Assignee: Stefania
> Fix For: 2.1.x
>
> Attachments: 6452.nps, 6452.png, 7300.nps, 7300a.png, 7300b.png,
> node1_debug.log, node2_debug.log, node3_debug.log, recording_127.0.0.1.jfr
>
>
> We get timeouts occasionally that cause the number of records to be incorrect:
> http://cassci.datastax.com/job/trunk_dtest/858/testReport/cqlsh_tests.cqlsh_copy_tests/CqlshCopyTest/test_bulk_round_trip_blogposts/
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
