[
https://issues.apache.org/jira/browse/CASSANALYTICS-138?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Josh McKenzie updated CASSANALYTICS-138:
----------------------------------------
Test and Documentation Plan: Run CI. See if it explodes.
Status: Patch Available (was: In Progress)
> Fix cluster startup retry logic in analytics dtests
> ---------------------------------------------------
>
> Key: CASSANALYTICS-138
> URL: https://issues.apache.org/jira/browse/CASSANALYTICS-138
> Project: Apache Cassandra Analytics
> Issue Type: Bug
> Components: Build
> Reporter: Josh McKenzie
> Assignee: Josh McKenzie
> Priority: Normal
> Fix For: 0.4
>
> Time Spent: 0.5h
> Remaining Estimate: 0h
>
> Seeing a few:
> {code:java}
> Caused by: org.apache.cassandra.exceptions.ConfigurationException:
> /127.0.0.1:35407 is in use by another process. Change
> listen_address:storage_port in cassandra.yaml to values that do not conflict
> with other services
> at
> org.apache.cassandra.net.InboundConnectionInitiator.bind(InboundConnectionInitiator.ja
> {code}
> when running tests. This is something we're supposed to gracefully handle
> w/retries in {{{}SharedClusterIntegrationTestBase.java{}}}, however it looks
> like that file both a) checks that the exception type is a
> {{{}BindException{}}}, and b) is hard coded to the singular text "Address
> already in use".
> On trunk in {{{}InboundConnectionInitiator.java{}}}, we both a) throw
> ConfigurationException, and b) have text reading "is in use by another
> process". We should remove specific exception type checking from our
> heuristic and just look for text containing either "Address already in use"
> or "is in use by another process" and have that be our trigger for graceful
> retries.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
