Bug 1301467

Summary: Dead L3 agent should reflect HA router states
Product: Red Hat OpenStack Reporter: Alexander Stafeyev <astafeye>
Component: openstack-neutronAssignee: Assaf Muller <amuller>
Status: CLOSED UPSTREAM QA Contact: Toni Freger <tfreger>
Severity: high Docs Contact:
Priority: low    
Version: 8.0 (Liberty)CC: amuller, chrisw, jschwarz, nyechiel, oblaut, srevivo, tfreger
Target Milestone: ---Keywords: ZStream
Target Release: ---   
Hardware: x86_64   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2017-09-07 22:03:20 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:

Description Alexander Stafeyev 2016-01-25 07:06:28 UTC
Bug tracker for -https://bugs.launchpad.net/neutron/+bug/1461148


l3 HA- when active is rebooted its state should be changed to "backup" and not stay active. 


Current- changes to backup only when it goes up again.

Comment 2 Assaf Muller 2016-06-01 00:49:56 UTC
All of the OSP 8 7.0.3+ builds should have the fix, I added an example build in the Fixed-In-Version field.

Comment 4 Toni Freger 2016-06-23 12:07:58 UTC
I've tested it in openstack-neutron-7.0.4-7.el7ost.noarch
I've followed step to reproduce, after the hosting active router booted the neutron l3-agent-list-hosting-router commands still shows 2 active agents:

| 31ebe946-7de7-4ea1-8e2b-895f97ac0c85 | rose11.qa.lab.tlv.redhat.com | True           | :-)   | active   |
| 070797e4-70a4-4137-ad76-3dcd60758315 | rose12.qa.lab.tlv.redhat.com | True           | :-)   | active

Comment 5 Assaf Muller 2016-06-23 15:01:01 UTC
(In reply to Toni Freger from comment #4)
> I've tested it in openstack-neutron-7.0.4-7.el7ost.noarch
> I've followed step to reproduce, after the hosting active router booted the
> neutron l3-agent-list-hosting-router commands still shows 2 active agents:
> 
> | 31ebe946-7de7-4ea1-8e2b-895f97ac0c85 | rose11.qa.lab.tlv.redhat.com | True
> | :-)   | active   |
> | 070797e4-70a4-4137-ad76-3dcd60758315 | rose12.qa.lab.tlv.redhat.com | True
> | :-)   | active

We use a default agent_down_time of 75 seconds. After which you should see 'xxx' on the agent and the ha_state should change to down.

Comment 10 Assaf Muller 2016-06-28 19:35:32 UTC
Assigned to John to look in to this when he comes in next week.

Comment 11 Assaf Muller 2017-01-13 19:13:31 UTC
Status update?

Comment 12 John Schwarz 2017-01-15 09:36:21 UTC
No update from me for this.

Comment 13 Assaf Muller 2017-09-07 22:03:20 UTC
I'm closing this bug, with a greater focus on OVN in a future OSP offering, allocating time and energy in low priority bugs that are specific to ML2/OVS doesn't make sense.

Comment 14 Ofer Blaut 2017-09-11 06:07:43 UTC
Fixed upstream - https://bugs.launchpad.net/neutron/+bug/1461148