[ 
https://issues.apache.org/jira/browse/SOLR-7615?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14568117#comment-14568117
 ] 

Mark Miller commented on SOLR-7615:
-----------------------------------

bq. that doesn't sound right at all. Are you sure the problem you're 
remembering having is the same code being discussed in this issue?

I specifically remember having to add OR's to 
LogUpdateProcessor.log.isInfoEnabled() to get that logging while using DEBUG 
logging.

Perhaps somehow I was smoking crack ... but I have very specific memories of 
making and caring this code change in a patch for a bit.

I'll take your word that the method behaves that way, but it doesn't fit a very 
specific memory of dealing with this code and using DEBUG logging. 

> Slow update requests not logged at WARN due to how LogUpdateProcessor's are 
> created
> -----------------------------------------------------------------------------------
>
>                 Key: SOLR-7615
>                 URL: https://issues.apache.org/jira/browse/SOLR-7615
>             Project: Solr
>          Issue Type: Bug
>            Reporter: Timothy Potter
>            Assignee: Timothy Potter
>            Priority: Minor
>
> In SOLR-6650, we introduced the ability to only have slow update and query 
> requests get logged as warnings. However, when tracking down the cause of 
> SOLR-7603, we discovered that the LogUpdateProcessor is only included in the 
> chain if the log has INFO enabled, thus rendering the solution provided by 
> SOLR-6650 useless for update requests (queries still work though).
> Thus, we need to fix how LogUpdateProcessor is created if the user has 
> configured slowUpdateThresholdMillis > 0.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org

Reply via email to