Docker client read timeout too low for certain calls
Bug #1450559 reported by
Andrew Melton
This bug affects 1 person
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Magnum |
Fix Released
|
Undecided
|
Unassigned |
Bug Description
Our Docker client is timing out certain calls before they are allowed to finish. Because Docker is a synchronous API it will block on calls until they complete. This means that certain calls can take quite a long time to return. For instance, when we make a container create call, Docker or Swarm is going to block and keep the connection open until it has pulled the image to and created the container on the host. For small images this isn't a problem, but on larger images, this can take a long time. So, we need a way to increase this timeout.
Example exception: http://
description: | updated |
Changed in magnum: | |
milestone: | none → mitaka-1 |
status: | Fix Committed → Fix Released |
To post a comment you must log in.
Reviewed: https:/ /review. openstack. org/179257 /git.openstack. org/cgit/ openstack/ magnum/ commit/ ?id=fa191a233a3 f53b4bd70a39d12 8bbe98f2b416a0
Committed: https:/
Submitter: Jenkins
Branch: master
commit fa191a233a3f53b 4bd70a39d128bbe 98f2b416a0
Author: Andrew Melton <email address hidden>
Date: Thu Apr 30 14:46:11 2015 -0700
Make Docker client timeout configurable
To support long running Docker operations we need to increase our
timeout, but a good value for the timeout will change per deployment.
So, add it as a config option so that it can be tweaked as necessary.
Change-Id: Ie53fa5086e7311 1bf9f42dbf59617 2af8d936cf0
Closes-Bug: #1450559