[
https://issues.apache.org/jira/browse/ARTEMIS-4207?focusedWorklogId=851395&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-851395
]
ASF GitHub Bot logged work on ARTEMIS-4207:
-------------------------------------------
Author: ASF GitHub Bot
Created on: 16/Mar/23 17:24
Start Date: 16/Mar/23 17:24
Worklog Time Spent: 10m
Work Description: gemmellr commented on code in PR #4404:
URL: https://github.com/apache/activemq-artemis/pull/4404#discussion_r1139104096
##########
artemis-server/src/main/java/org/apache/activemq/artemis/core/postoffice/impl/PostOfficeImpl.java:
##########
@@ -1391,7 +1395,8 @@ public Pair<RoutingContext, Message> redistribute(final
Message message,
// We have to copy the message and store it separately, otherwise we
may lose remote bindings in case of restart before the message
// arrived the target node
// as described on https://issues.jboss.org/browse/JBPAPP-6130
- Message copyRedistribute = message.copy(storageManager.generateID());
+ final Message copyRedistribute =
message.copy(storageManager.generateID());
+ logger.info("Message {} being copied as {}", message.getMessageID(),
copyRedistribute.getMessageID());
Review Comment:
I find it a good habit to mark such 'temp dev/debugging logging' changes
with TODOs when making the change, its then much harder to forget to miss e.g
setting them to the correct level (or removing them) later when reviewing
things before pushing.
Issue Time Tracking
-------------------
Worklog Id: (was: 851395)
Time Spent: 40m (was: 0.5h)
> Redistribution may leave large messages stranded
> ------------------------------------------------
>
> Key: ARTEMIS-4207
> URL: https://issues.apache.org/jira/browse/ARTEMIS-4207
> Project: ActiveMQ Artemis
> Issue Type: Bug
> Components: Broker
> Affects Versions: 2.28.0
> Reporter: Clebert Suconic
> Priority: Major
> Fix For: 2.29.0
>
> Time Spent: 40m
> Remaining Estimate: 0h
>
> If you are redistributing messages, eventually between the copy of the large
> message and the actual call to redistribute the routing table may change,
> leaving a message behind.
> I could see this happening more oftenly on a test where I had 2 tests in
> clustered, and one server was shutdown with Ctrl-C (SIGINT) rather than kill
> -9.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
