Note: This bug is displayed in read-only format because the product is no longer active in Red Hat Bugzilla.

Bug 1520673

Summary:	User cannot create additional routers in a tenant after deleting a router, if using HA routers
Product:	Red Hat OpenStack	Reporter:	Assaf Muller <amuller>
Component:	openstack-neutron	Assignee:	Brian Haley <bhaley>
Status:	CLOSED ERRATA	QA Contact:	Roee Agiman <ragiman>
Severity:	urgent	Docs Contact:
Priority:	urgent
Version:	12.0 (Pike)	CC:	amuller, chrisw, jlibosva, jschluet, lpeer, markmc, nyechiel, oblaut, sclewis, srevivo, tvignaud
Target Milestone:	ga	Keywords:	Regression, Triaged, ZStream
Target Release:	12.0 (Pike)
Hardware:	Unspecified
OS:	Unspecified
Whiteboard:
Fixed In Version:	openstack-neutron-11.0.1-7.el7ost	Doc Type:	If docs needed, set a value
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2017-12-13 19:41:31 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:
Embargoed:

Description Assaf Muller 2017-12-04 22:28:24 UTC

Steps to Reproduce:
1. On a fresh tenant with zero HA routers and no HA network, create an HA router
2. Create a second HA router
3. Delete one of the routers, this will keep the tenant's HA network around but delete its mapping to a segmentation ID (VXLAN VNI, GRE tunnel ID or VLAN)
4. It's now impossible to create more HA routers
5. Restarting a L3 agent will at this point also break the remaining router
6. The only workaround at this point is to delete all HA routers and create them again, resulting in data plane downtime.

The Launchpad bug has more details, in particular comment 8 (https://bugs.launchpad.net/neutron/+bug/1732543/comments/8).

Comment 1 Assaf Muller 2017-12-04 22:31:48 UTC

Note: This is a regression from OSP 11.

Comment 2 Assaf Muller 2017-12-04 22:34:29 UTC

Brian and I spoke on IRC, he reports this was introduced in change https://review.openstack.org/#/c/423584/.

Comment 5 Brian Haley 2017-12-05 15:17:53 UTC

I would hope to have an upstream patch out soon today for review, at best it would merge tonight.  Realistically that probably means tomorrow, but it's my top priority right now.

Comment 6 Mark McLoughlin 2017-12-06 05:38:54 UTC

It looks like this is the upstream patch under review for this bug: https://review.openstack.org/475955

Comment 7 Brian Haley 2017-12-06 05:56:42 UTC

Yes, that should fix the issue and would be an easy backport to stable/pike, it just needs some more testing on my part.

There will be a follow-on to truly fix this code path, as the way the notifiers are being used isn't correct, but that's not causing a critical issue so can wait until after it merges.

Comment 8 Assaf Muller 2017-12-06 13:27:53 UTC

This is also being tested as we speak (in parallel to upstream development) by the DFG's CI and QE specialists.

Comment 20 Brian Haley 2017-12-11 15:21:25 UTC

Upstream tempest test:

https://review.openstack.org/#/c/526134

Comment 23 errata-xmlrpc 2017-12-13 19:41:31 UTC

Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory, and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://access.redhat.com/errata/RHBA-2017:3461