Moving away from ActiveMQ

AcitveMQ was meant to be a central part of our infrastructure. To access it from Python we where using the Stomp Protocol.

Unfortunately while testing our Stomp Protocol implementation there where occasional hickups during the unittests. Nasty network deadlocks? Obscure non-blocking socket issues? Messed up state machine?

I tested a different implementation and I saw similar or even worse issues.

Could this issues root in the Broker and not in the Protocol implementation on the client site? The Apache branding in Active MQ, the Version Number 4.1 and the prominent advertising for the Stomp Protocol suggests a mature application and a vivid community. Then again prospective bug reports don’t get answered, bugs get closed with „is no bug since something else works“ and there are many unfixed bugs related to Stomp.

So my conclusion is: ActiveMQ & Stomp aren’t ready for serious, high-load usage so far. A Pitty that I wasted several days finding that out.

2 comments on “Moving away from ActiveMQ

  1. aksyn
    2008-10-16 at 00:07 #

    RabbitMQ

    RabbitMQ
    Have you tried RabbitMQ (http://www.rabbitmq.com/) ?

    This comment was originally posted on 20070829T18:33:32

  2. mdornseif
    2008-10-16 at 00:07 #

    Yes. But there are/where issues with the Erlang and Python Bindings.

    This comment was originally posted on 20071207T07:17:58

Kommentar verfassen

Trage deine Daten unten ein oder klicke ein Icon um dich einzuloggen:

WordPress.com-Logo

Du kommentierst mit Deinem WordPress.com-Konto. Abmelden / Ändern )

Twitter-Bild

Du kommentierst mit Deinem Twitter-Konto. Abmelden / Ändern )

Facebook-Foto

Du kommentierst mit Deinem Facebook-Konto. Abmelden / Ändern )

Google+ Foto

Du kommentierst mit Deinem Google+-Konto. Abmelden / Ändern )

Verbinde mit %s