[
https://issues.apache.org/jira/browse/SOLR-7344?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14585216#comment-14585216
]
Yonik Seeley commented on SOLR-7344:
------------------------------------
bq. It should work with whatever limits - you should be able to specify
whatever sizes and avoid deadlock.
But what's the plan to make it work with whatever sizes?
bq. The default is not that important
I don't care about the tweaks (changing from 32 to 64 because they have a
bigger box), I care about one queue size being an order of magnitude bigger
than another queue size for some good reason, or which queue sizes need to be
(essentially) unbounded to prevent distributed deadlock. Defaults can make the
difference between deadlock and no deadlock.
bq. Behavior over the limit is to start rejecting requests....
Ouch... I hope that's not true. That would certainly be very invasive to
clients. Hopefully there's blocking behavior in there too, and requests are
only rejected after too many of them are buffered?
> Allow Jetty thread pool limits while still avoiding distributed deadlock.
> -------------------------------------------------------------------------
>
> Key: SOLR-7344
> URL: https://issues.apache.org/jira/browse/SOLR-7344
> Project: Solr
> Issue Type: Improvement
> Components: SolrCloud
> Reporter: Mark Miller
> Attachments: SOLR-7344.patch
>
>
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
