Uploaded image for project: 'Red Hat OpenStack Services on OpenShift'
  1. Red Hat OpenStack Services on OpenShift
  2. OSPRH-19901

[rabbitmq 4.1.4] Implement cluster exchange limit detection logic.

XMLWordPrintable

    • 5
    • False
    • Hide

      None

      Show
      None
    • False
    • ?
    • rhos-ops-platform-services-pidone
    • None
    • Sprint 4, Sprint 7
    • 2
    • Moderate

      To Reproduce{}

      Since version 4.1.4 RabbitMQ allow to configure a limit [1][2] on the cluster-wide number of exchanges that applications.

      This new RabbitMQ feature is a guard against applications that leak resources. However oslo.messaging as a client lib is not prepared in case operators decide to setup this limit.

      [1] https://github.com/rabbitmq/rabbitmq-server/releases/tag/v4.1.4
      [2] https://github.com/rabbitmq/rabbitmq-server/pull/14339/files

      Expected behavior

      Oslo.messaging should raise a controlled exception in case the cluster refuse to create a new exchange if this limit is reached.

      This way service would be able to implement retry logic on their side

      Bug impact

      As this error is not handled services may break if operator setup the cluster limit. 

      •  

              hberaud Hervé Beraud
              hberaud Hervé Beraud
              rhos-dfg-pidone
              Votes:
              0 Vote for this issue
              Watchers:
              1 Start watching this issue

                Created:
                Updated: