Retrying to AMQP server is going beyond the value of rabbit_interval_max

[abdul nizamuddin]

When rabbitmq is down and if the user run any heat commands and user is not getting any errors messages.
In logs i can see for every 2seconds it keeps retrying to connect to AMQP server. And this retry reaches till 32 seconds.After the 32 seconds it wil retry again to connect to AMQP server and fails and again after 32 seconds it tries to connect and for every 32 seconds it tries to connect.
But in heat.conf i have set the value for the parameter rabbit_interval_max=30. this value is default value. This retry interval should not exceed more 30 retrries.

[Ken Giusti (kgiusti)]

I can reproduce this outside of oslo.messaging using the latest kombu directly - see attached test.

So this is not a bug in oslo.messaging but in the kombu library that oslo.messaging uses.

I have opened a bug on celery for this:

https://github.com/celery/kombu/issues/914

[Ken Giusti (kgiusti)]

Marking it won't fix since hacking a work around is probably more risk than it's worth since we have to write our own custom version of autoretry()