[
https://issues.apache.org/jira/browse/SOLR-12766?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16613666#comment-16613666
]
Tomás Fernández Löbbe commented on SOLR-12766:
----------------------------------------------
Ah, I forgot that for TOLEADER requests we retry 25 times, and I made the
backoff linear without cap... That's a very slow retry. I'll make the backoff:
{code:java}
int backoffTime = Math.min(retryPause * resubmitList.get(0).req.retries, 2000);
{code}
> When retrying internal requests, backoff only once for the full batch of
> retries
> --------------------------------------------------------------------------------
>
> Key: SOLR-12766
> URL: https://issues.apache.org/jira/browse/SOLR-12766
> Project: Solr
> Issue Type: Improvement
> Security Level: Public(Default Security Level. Issues are Public)
> Reporter: Tomás Fernández Löbbe
> Assignee: Tomás Fernández Löbbe
> Priority: Minor
> Attachments: SOLR-12766.patch
>
>
> We currently wait for each internal retry request ({{TOLEADER}} or
> {{FROMLEADER}} requests). This can cause a long wait time when retrying many
> requests and can timeout the client. We should instead wait once and retry
> the full batch of errors.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
