Nameko behavior on rabbitmq crash


#1

What will happen if the rabbitmq crashes, will nameko reconnect after rabbitmq comes up?.
we are seeing that with only 4-5 services and low load i.e. 4 msg/sec , rabbitmq inside a docker , is crashing and when rabbitmq crashes, all services stops or get hung

i.e. we are seeing behavior similar to this https://groups.google.com/forum/?nomobile=true#!msg/nameko-dev/POBxAV71TUE/XRqyV-iHCwAJ


#2

What will happen if the rabbitmq crashes, will nameko reconnect after rabbitmq comes up?

Yes. As long as the process stays up, the consumers should attempt to reconnect.

we are seeing behavior similar to this https://groups.google.com/forum/?nomobile=true#!msg/nameko-dev/POBxAV71TUE/XRqyV-iHCwAJ

Hanging and crashing would mean very different causes. If you see the process terminate with a crash, you may be experiencing the above. It very specifically happens when an exception occurs inside get_dependency. This is a “feature” that we could easily remove, and is also easily worked-around by running Nameko under a process manager like supervisord.

Please post a stacktrace or otherwise describe the “hang” in more detail.