You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@activemq.apache.org by panu <pa...@team.telstra.com> on 2007/08/16 08:45:11 UTC

A few issues with Stomp

Hi Guys,

    I just want to report two issues I have experienced while using
ActiveMQ. I use Stomp (my own Python implementation) as the protocol to
communicate with ActiveMQ. 

    - Messages are stuck in the queue when multiple consumers process series
of messages in the second time. In this case, I have two consumers consuming
messages from the same queue. I feed about 5000 messages of size 1K into the
queue, things are fine. I again feed the same 5000 messages, but this time
these messages are stuck in the queue. I check the consumers; they wait for
the messages on socket recv() with no error. It seems to me that ActiveMQ
does not forward the messages to the consumers. When the consumers reconnect
(I add timeout just to workaround this problem), they then get these 5000
messages.

    - The consumer of a Message Group is changed when JMSXGroupSeq is set to
1 not 0. Based on the document I read from the website, the consumer for a
group of messages is suppopsed to change when JMSXGroupSeq is set to 0.
However, I have tested this many times and found that only when I set
JMSXGroupSeq to 1, the consumer for that group of messages is changed, which
I also get JMSXGroupFirstForConsumer = true. Also, sometimes, the same
consumer is chosen instead of a new consumer.

    Normally, I use SNAPSHOT version 4.2-20070513.230550-57, but I also
tested with SNAPSHOT version 5.0-20070815.220536-52, and I found these
issues in both versions.

Regards,
Panu
-- 
View this message in context: http://www.nabble.com/A-few-issues-with-Stomp-tf4277814s2354.html#a12176133
Sent from the ActiveMQ - User mailing list archive at Nabble.com.