(Migrated) Limit and hang when processing incoming AMQP messages

(This message has been automatically imported from the retired mailing list)


I have encountered an interesting problem over the last few days. On my
ZATO instance I have an incoming AMQP channel that is linked to a
service whose only job is to store the message in a Mongo DB.

For various reasons the processing of messages has fallen behind and I
now have approximately 181,000 message in the queue waiting to be processed.

ZATO does not appear to be processing these messages. I am seeing the
following behaviour:

  • no error messages on startup,
  • after starting I can see 400 messages in the AMQP queue (rabbitMQ)
    change from Ready to Unacknowledged,
  • I can see around 250K of data in the TCP Recv-Q of one of the
    channel.py processes,
  • all other inbound AMQP messages are being processed immediately with
    no problems.

Any suggestions?

ZATO version is 2.0.3.