[
https://issues.apache.org/jira/browse/SOLR-6405?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14107325#comment-14107325
]
Mark Miller commented on SOLR-6405:
-----------------------------------
bq. we can really just retry "forever" without limiting the loop count
You can get other exceptions as well though. The original implementation around
this waited forever I think and it caused a lot of issues in some cases -
perhaps that was mainly in failure cases. I don't really remember the details
currently and a lot has changed since then.
bq. The advantage of this approach is to never sleep for too long before
finding out the definitive answer of success or SessionExpired
Yeah, I guess it's just not been a big concern of mine since the extra wait is
not really enough to matter, and the worst cases is not common or that bad
either. This should be an exceptional case and either it's a quick blip or a
really expensive recovery anyway. I didn't really see it as a goal to get out
as fast as possible because there didn't seem to be much benefit, just to
ensure we waited longer than expiration.
If you want to explore this alternative approach, let's open a new JIRA for it.
> ZooKeeper calls can easily not be retried enough on ConnectionLoss.
> -------------------------------------------------------------------
>
> Key: SOLR-6405
> URL: https://issues.apache.org/jira/browse/SOLR-6405
> Project: Solr
> Issue Type: Bug
> Components: SolrCloud
> Reporter: Mark Miller
> Assignee: Mark Miller
> Priority: Critical
> Fix For: 5.0, 4.10
>
> Attachments: SOLR-6405.patch
>
>
> The current design requires that we are sure we retry on connection loss
> until session expiration.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org
