[
https://issues.apache.org/jira/browse/SOLR-18422?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18113943#comment-18113943
]
ASF subversion and git services commented on SOLR-18422:
--------------------------------------------------------
Commit ccb0d010efc6defb4e77bb2cd842be2a32fca710 in solr's branch
refs/heads/main from Andrzej Bialecki
[ https://gitbox.apache.org/repos/asf?p=solr.git;h=ccb0d010efc ]
SOLR-18422: Don't use System.nanoTime as absolute timestamp in CrossDC
Producer. (#4888)
> CrossDC Producer: System.nanoTime used as absolute timestamp
> ------------------------------------------------------------
>
> Key: SOLR-18422
> URL: https://issues.apache.org/jira/browse/SOLR-18422
> Project: Solr
> Issue Type: Bug
> Components: module - crossDC
> Affects Versions: 10.0, main(11.0), 9.10.1
> Reporter: Andrzej Bialecki
> Assignee: Andrzej Bialecki
> Priority: Minor
> Labels: pull-request-available
> Time Spent: 20m
> Remaining Estimate: 0h
>
> {{SolrMessageProcessor.logFirstAttemptLatency}} logs and records a metric
> that is currently meaningless because it treats the
> {{MirroredSolrRequest.getSubmitTimeNanos()}} as an absolute timestamp - but
> this value is initialized on the producer side from {{System.nanoTime()}}
> which has an arbitrary origin.
> Even assuming that both sides of the CrossDC path are perfectly
> time-synchronized currently this value is meaningless. Instead, both sides
> should use absolute timestamps from {{TimeSource.CURRENT_TIME.getTimeNs()}}
> to at least have a chance of producing values closer to reality.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
