[
https://issues.apache.org/jira/browse/SOLR-9367?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15414295#comment-15414295
]
ASF subversion and git services commented on SOLR-9367:
-------------------------------------------------------
Commit 84a8c098fca2018ca8c6abde13ae0d56f081dae1 in lucene-solr's branch
refs/heads/master from Chris Hostetter
[ https://git-wip-us.apache.org/repos/asf?p=lucene-solr.git;h=84a8c09 ]
SOLR-9367: Improved TestInjection's randomization logic to use
LuceneTestCase.random()
> TestInjection's use of a static Random means that sequences of events can be
> diff for the same test seed depending on wether individual methods are run,
> or entire test classes, or multiple classes in the same slave JVM
> --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
>
> Key: SOLR-9367
> URL: https://issues.apache.org/jira/browse/SOLR-9367
> Project: Solr
> Issue Type: Bug
> Security Level: Public(Default Security Level. Issues are Public)
> Reporter: Hoss Man
> Fix For: 6.2, master (7.0)
>
> Attachments: SOLR-9367.patch
>
>
> While digging through SOLR-9363, I realized that the reason the failures
> reproduced when running the entire {{-Dtestcase}} but not the individual
> {{-Dtests.method}} was because TestInjection was behaving differently in
> those cases.
> This is because TestInjection uses a static Random instance -- allthough it
> is initialized based on the {{tests.seed}} sysprop to _try_ to be
> reproducible, it's behavior still very much depends on how many testcases are
> run in that slave JVM, and/or how many test methods are run.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
