Page MenuHomePhabricator

coe timeout
Closed, ResolvedPublic

Description

I'm getting the following when trying to run coe commands from my laptop:
Request to https://openstack.eqiad1.wikimediacloud.org:29511/v1/clusters timed out

which seems reasonable, as that endpoint doesn't seem to exist on openstack with the /clusters suffix.

I've installed the python3-magnumclient (I'm running ubuntu 22.04.1) which appears to get me much the same version as exists on cloudcontrol1005. Mine:

ii  python3-magnumclient                  3.5.0-0ubuntu3            all          client library for Magnum API - Python 3.x

cloudcontrol1005:

ii  python3-magnumclient                  3.5.0-3~bpo11+1           all          client library for Magnum API - Python 3.x

coe commands do seem to work fine from cloudcontrol1005 itself.

I'm guessing this is either my client is trying the wrong endpoint (seems to work with other things like openstack server list) or something isn't being allowed from external sources?

Related Objects

StatusSubtypeAssignedTask
Resolved rook
ResolvedNone

Event Timeline

Change 844457 had a related patch set uploaded (by Majavah; author: Majavah):

[operations/puppet@production] P:openstack: expose remaining APIs to the internet

https://gerrit.wikimedia.org/r/844457

Local solution found (export OS_PROJECT_ID=testlabs) for testing this is fine.

Change 844457 abandoned by Majavah:

[operations/puppet@production] P:openstack: expose remaining APIs to the internet

Reason:

superseded by cloudlb work

https://gerrit.wikimedia.org/r/844457