[
https://issues.apache.org/jira/browse/CASSANDRA-10167?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14715421#comment-14715421
]
Jim Witschey edited comment on CASSANDRA-10167 at 8/26/15 8:18 PM:
-------------------------------------------------------------------
I'm not sure what causes the failure, but as far as I can tell, here's what's
happening in the test:
- start a 2-node cluster
- load data with stress
- flush node 1, then start a job in a separate thread that will kill it during
node 3's bootstrapping
- add and start node 3.
Sometimes, that third node fails to start, so the test times out waiting to see
the 3rd node start in its log.
EDIT: was wrong about how node 1 is killed above.
was (Author: mambocab):
I'm not sure what causes the failure, but as far as I can tell, here's what's
happening in the test:
- start a 2-node cluster
- load data with stress
- flush node1, then kill it while it bootstraps
- add and start a third node.
Sometimes, that third node fails to start, and the the test times out waiting
to see the 3rd node start in its log.
> fix resumable_bootstrap_test dtest failures
> -------------------------------------------
>
> Key: CASSANDRA-10167
> URL: https://issues.apache.org/jira/browse/CASSANDRA-10167
> Project: Cassandra
> Issue Type: Bug
> Reporter: Jim Witschey
> Assignee: Yuki Morishita
>
> {{resumable_bootstrap_test}} fails periodically and I believe that's been
> true for quite a while. This is related to CASSANDRA-9132 and CASSANDRA-9444.
> the current failures look like this:
> http://cassci.datastax.com/job/cassandra-2.2_dtest/228/testReport/junit/bootstrap_test/TestBootstrap/resumable_bootstrap_test/
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
