[ 
https://issues.apache.org/jira/browse/AMQ-9255?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17763151#comment-17763151
 ] 

Nicolas Filotto commented on AMQ-9255:
--------------------------------------

A potential fix https://github.com/apache/activemq/pull/1051

> Messages submitted via http(s) transport don't dead letter after TTL is 
> exceeded
> --------------------------------------------------------------------------------
>
>                 Key: AMQ-9255
>                 URL: https://issues.apache.org/jira/browse/AMQ-9255
>             Project: ActiveMQ
>          Issue Type: Bug
>          Components: JMS client
>    Affects Versions: 5.16.6
>         Environment: Tested on 5.16.6, Windows client, Windows & Linux 
> servers. 
> openjdk version "1.8.0_322"
> OpenJDK Runtime Environment (Temurin)(build 1.8.0_322-b06)
>            Reporter: Paul McCulloch
>            Priority: Major
>          Time Spent: 10m
>  Remaining Estimate: 0h
>
> When a persistent messages is submitted to a queue, with a time to live 
> specified, it should be removed from the queue and (in the default 
> configuration) moved to the DLQ. 
> This is the case with the web console, and with a trivial Java client using 
> the openwire transport (specified using a "tcp://" broker URL when creating 
> ActiveMQConnectionFactory.
> If an http/https url is specified this doesn't behave correctly. When the TTL 
> is exceeded the message seems to go into a limbo state. It doesn't get 
> delivered to consumers, but also doesn't go to the DLQ (as determined via the 
> web console).
> The server logs an error when using the web console (stack trace at end):
> java.lang.NullPointerException
>     at 
> org.apache.activemq.command.Message.canProcessAsExpired(Message.java:865)
> I believe the cause of the web console issue (and perhaps the whole DLQ 
> thing) is that a transient field, processAsExpired, was added to 
> org.apache.activemq.command.Message, but Message doesn't implement 
> TransientInitializer.
> The OpenWire transport doesn't suffer from this as its command unmarshaling 
> code uses a constructor to initialize the ActiveMQTextMessage  - so the 
> transient field isn't null.
>  
> Stack Trace:
> 2023-04-30 06:38:00,583 | ERROR | org.apache.activemq.broker.region.Queue | 
> Problem retrieving message for browse | ActiveMQ BrokerService[localhost] 
> Task-2
> java.lang.NullPointerException
>     at 
> org.apache.activemq.command.Message.canProcessAsExpired(Message.java:865)
>     at 
> org.apache.activemq.broker.region.IndirectMessageReference.canProcessAsExpired(IndirectMessageReference.java:210)
>     at 
> org.apache.activemq.broker.region.RegionBroker.isExpired(RegionBroker.java:749)
>     at 
> org.apache.activemq.broker.BrokerFilter.isExpired(BrokerFilter.java:309)
>     at 
> org.apache.activemq.broker.BrokerFilter.isExpired(BrokerFilter.java:309)
>     at 
> org.apache.activemq.broker.BrokerFilter.isExpired(BrokerFilter.java:309)
>     at 
> org.apache.activemq.broker.BrokerFilter.isExpired(BrokerFilter.java:309)
>     at 
> org.apache.activemq.broker.BrokerFilter.isExpired(BrokerFilter.java:309)
>     at org.apache.activemq.broker.region.Queue.doBrowseList(Queue.java:1216)
>     at org.apache.activemq.broker.region.Queue.doBrowse(Queue.java:1196)
>     at org.apache.activemq.broker.region.Queue.expireMessages(Queue.java:964)
>     at org.apache.activemq.broker.region.Queue.access$100(Queue.java:108)
>     at org.apache.activemq.broker.region.Queue$2.run(Queue.java:152)
>     at 
> java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
>     at 
> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
>     at java.lang.Thread.run(Thread.java:750)
>  



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

Reply via email to