[
https://issues.apache.org/jira/browse/NIFI-8353?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17307323#comment-17307323
]
ASF subversion and git services commented on NIFI-8353:
-------------------------------------------------------
Commit 321e979673139310b7f92dfccd926092a2b6c1c9 in nifi's branch
refs/heads/main from markap14
[ https://gitbox.apache.org/repos/asf?p=nifi.git;h=321e979 ]
NIFI-8353: When receiving data via load-balanced connection, throw an Exception
(resulting in an ABORT_TRANSACTION status code) when attempting to add received
FlowFiles to the FlowFile queue, if the node is not currently connected to
cluster. In this case, ensure that we remove the received FlowFiles from the
FlowFile Repository and emit a DROP event to the repository (with appropriate
details) to coincide with the already-committed Provenance events. Also ensure
that when a node is disconnected that we don't keep sending to that node until
it reconnects. While testing the fixes via LoadBalancedQueueIT, also noticed
that unit test logging was not working properly in nifi-framework-core because
of the (erroneous) dependency on logback-class; removed this dependency and
updated unit tests that were created that depended on its existence. (#4924)
> When node is offloaded, it may still receive data from load-balanced
> connections
> --------------------------------------------------------------------------------
>
> Key: NIFI-8353
> URL: https://issues.apache.org/jira/browse/NIFI-8353
> Project: Apache NiFi
> Issue Type: Bug
> Components: Core Framework
> Reporter: Mark Payne
> Assignee: Mark Payne
> Priority: Blocker
> Fix For: 1.14.0
>
> Time Spent: 1h
> Remaining Estimate: 0h
>
> Consider a simple flow: GenerateFlowFile (9 FlowFiles per batch, each 10 KB,
> scheduled once per second) -> UpdateAttribute.
> The connection between the two processors is configured with Round Robin load
> balancing.
> Start GenerateFlowFile. Wait a few seconds for some data to queue up.
> Disconnect Node 2. Immediately offload Node 2. Refresh cluster page to see
> that the node is offloaded.
> Navigate to Node 2's UI. Sometimes it will show data queued up. This is
> because after the node was offloaded, the connection from Node 1 was still
> transferring data to Node 2. Node 1 won't start any additional transactions
> to send data to Node 2, but it will complete the already-started
> transactions. So the node can be offloaded and end up with data queued.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
