[
https://issues.apache.org/jira/browse/METRON-820?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15953957#comment-15953957
]
ASF GitHub Bot commented on METRON-820:
---------------------------------------
Github user cestella commented on the issue:
https://github.com/apache/incubator-metron/pull/508
@ottobackwards It should not. When we start being able to handle people
dropping new jars in HDFS on a *running* topology, then we will need to prune
this cache. As of now, though, a new jar in HDFS requires a topology restart,
so this cache would always be fresh.
> StellarProcessor should have a static expression cache
> ------------------------------------------------------
>
> Key: METRON-820
> URL: https://issues.apache.org/jira/browse/METRON-820
> Project: Metron
> Issue Type: Bug
> Reporter: Casey Stella
>
> We recreate StellarProcessor objects all over the place which do not reuse
> their expression caches. This results in a significant performance
> regression as we are compiling the expression and then processing for every
> message, rather than compiling once and reusing.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
