OpenStack Nova Cloud Controller Charm

Check_haproxy_servers reports OK if only 1 port of a backend is reachable

Bug #1811397 reported by Boroczki, Lajos
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
OpenStack Nova Cloud Controller Charm
Triaged
Medium
Unassigned

Bug Description

Our nova-cloud-controller setup has e.g. nova-api-ec2 disabled, and it is configured in haproxy.cfg (Bug: https://bugs.launchpad.net/charm-nova-cloud-controller/+bug/1811398)

Check_haproxy_servers returns OK in this case, which is invalid; because all of the backends of nova-api-ec2 frontend are down.

Another issue with current check_haproxy_servers script is that it executes check_http against the same url (with a timeout of 10 seconds) about 16 times, while nagios default timeout is also 10 seconds.

I'm attaching the proposed check_haproxy.sh script.

See original description
Revision history for this message
Boroczki, Lajos (boroczki-lajos) wrote :
description: updated
Chris MacNaughton (chris.macnaughton)
Changed in charm-nova-cloud-controller:
status: New → Triaged
importance: Undecided → Medium
Revision history for this message
Gábor Mészáros (gabor.meszaros) wrote :

Lajos, the generic way of proposing a fix is not via attaching a file on LP, but creating a review on the Openstack site, as charms are part of the official Openstack review process.
Could you open a review there and link it here? Thanks

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.
Subscribing...

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.