Comment 7 for bug 1521958

Reviewed: https://review.openstack.org/252359
Committed: https://git.openstack.org/cgit/openstack/oslo.messaging/commit/?id=8504e2508bbec963ae817cc31fec509d058c0e96
Submitter: Jenkins
Branch: stable/liberty

commit 8504e2508bbec963ae817cc31fec509d058c0e96
Author: Mehdi Abaakouk <email address hidden>
Date: Wed Dec 2 11:38:27 2015 +0100

    Don't hold the connection when reply fail

    This change moves the reply retry code to upper layer
    to be able to release the connection while we wait between
    two retries.

    In the worse scenario, a client waits for more than 30 replies
    and died/restart, the server tries to send this 30 replies to this
    this client and can wait too 60s per replies. During this
    replies for other clients are just stuck.

    This change fixes that.

    Related-bug: #1477914
    Closes-bug: #1521958

    (cherry picked from commit I0d3c16ea6d2c1da143de4924b3be41d1cea159bd)

    Conflicts:
     oslo_messaging/_drivers/amqpdriver.py
     oslo_messaging/_drivers/impl_rabbit.py

    Depends-On: Ibce834c3e76d71a770013cf1b469aa86396751b9
    Change-Id: I18144ede387e1d28f7b5de0131b6b6cc7d57bb86