[
https://issues.apache.org/jira/browse/NIFI-16476?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18129578#comment-18129578
]
ASF subversion and git services commented on NIFI-16476:
--------------------------------------------------------
Commit 93caeedd42a751e333089e28cd9040aa9316671e in nifi's branch
refs/heads/main from boguszj
[ https://gitbox.apache.org/repos/asf?p=nifi.git;h=93caeedd42a ]
NIFI-16476 Fix last queue dates for queues with many empty remote partitions
and priority queues (#11821)
> Connection max queued duration reports 0 for clustered connections with empty
> partitions
> ----------------------------------------------------------------------------------------
>
> Key: NIFI-16476
> URL: https://issues.apache.org/jira/browse/NIFI-16476
> Project: Apache NiFi
> Issue Type: Bug
> Components: Core Framework
> Reporter: Jakub Bogusz
> Assignee: Jakub Bogusz
> Priority: Minor
> Time Spent: 20m
> Remaining Estimate: 0h
>
> On clustered nodes, connection max queued duration is often reported as 0
> even though FlowFiles are queued.
> SocketLoadBalancedFlowFileQueue.getMinLastQueueDate() treats the 0 returned
> by empty partitions as the oldest queue date. Non-load-balanced connections
> always have empty remote partitions, so they are affected most.
> SwappablePriorityQueue.getMinLastQueueDate() has the same flaw when the
> active queue is empty but the swap queue is not.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
