Container logs operation hangs until timeout

[Andrew Melton]

It appears we are using the old style method to get logs where dockerpy attaches to the container and attempts to stream the logs. That appears to hang with our current set up. This doesn't appear to happen with the newer 'logs' operation exposed by dockerpy.

Steps to reproduce:
1) Spawn Swarm Bay
2) Create a container
3) Run magnum container-start <container>
4) Run magnum container-logs <container>
5) Magnum client should hang until it gets a timeout exception from Magnum

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/229627

[Egor Guz (eghobo)]

@Andrew: just curious is it's atomic 2 or 5. Keep in mind that 5 has Docker 1.7.1

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/229627
Committed: https://git.openstack.org/cgit/openstack/magnum/commit/?id=63cd7f4017aeff5eac9fe5c8713ec27954c58d09
Submitter: Jenkins
Branch: master

commit 63cd7f4017aeff5eac9fe5c8713ec27954c58d09
Author: Andrew Melton <email address hidden>
Date: Wed Sep 30 16:44:40 2015 -0400

    Use dockerpy logs operation instead of attach

    Attach is the old way of retrieving container logs, and it hangs
    when used with Swarm. Dockerpy handles the code to switch between
    the old attach and the new logs operation depending on remote
    version, so this should be a backwards compatible change.

    Change-Id: I224ee41c634a2c7cc1aa63e0495668063f66db80
    Closes-Bug: #1501498