| 2015-03-17 16:00:20 |
Assaf Muller |
bug |
|
|
added bug |
| 2015-03-17 16:00:27 |
Assaf Muller |
neutron: assignee |
|
Assaf Muller (amuller) |
|
| 2015-03-17 16:00:31 |
Assaf Muller |
neutron: status |
New |
In Progress |
|
| 2015-03-17 18:21:53 |
Sridhar Gaddam |
bug |
|
|
added subscriber Sridhar Gaddam |
| 2015-04-30 17:48:50 |
Assaf Muller |
neutron: status |
In Progress |
Triaged |
|
| 2015-04-30 17:48:54 |
Assaf Muller |
neutron: importance |
Undecided |
Medium |
|
| 2015-04-30 17:49:41 |
Assaf Muller |
neutron: importance |
Medium |
Low |
|
| 2015-04-30 23:09:35 |
Assaf Muller |
description |
keepalived 1.2.14 introduced a regression when running it in no-preempt mode. More details here in a thread I started on the keepalived-devel list:
http://sourceforge.net/p/keepalived/mailman/message/33604497/
A fix will possibly be sent and packaged in 1.2.16. Maybe it will be backported to 1.2.14 and 1.2.15. We now have two choices:
1) Find a work around in Neutron. The only one I could find is to enable pre-emption, which has its set of pros and cons.
2) Wait until a fix lands, 1.2.16 is released and work with the distros to backport it to 1.2.14 and 1.2.15. Any deployments that are running the affected versions are screwed in the mean time. |
keepalived 1.2.14 introduced a regression when running it in no-preempt mode. More details here in a thread I started on the keepalived-devel list:
http://sourceforge.net/p/keepalived/mailman/message/33604497/
A fix was backported to 1.2.15-6, and is present in 1.2.16.
Current status (Updated on the 30th of April, 2015):
Fedora 20, 21 and 22 have 1.2.16.
CentOS and RHEL are on 1.2.13
Ubuntu is using 1.2.10 or older.
Debian is using 1.2.13.
In summary, as long as you're not using 1.2.14 or 1.2.15 (Excluding 1.2.15-6), you're OK, which should be the case if you're using the latest keepalived packaged for your distro. |
|
| 2015-12-04 22:05:35 |
Assaf Muller |
neutron: assignee |
Assaf Muller (amuller) |
|
|
| 2015-12-04 22:06:09 |
Assaf Muller |
summary |
L3 HA routers master state flapping between nodes after router updates or failovers |
L3 HA routers master state flapping between nodes after router updates or failovers when using 1.2.14 or 1.2.15 (-1.2.15-6) |
|
| 2016-07-01 09:57:52 |
Jesse Pretorius |
bug task added |
|
openstack-ansible |
|
| 2016-07-05 16:59:31 |
Jean-Philippe Evrard |
openstack-ansible: assignee |
|
Jean-Philippe Evrard (jean-philippe-evrard) |
|
| 2016-07-05 16:59:38 |
Jean-Philippe Evrard |
openstack-ansible: status |
New |
Confirmed |
|
| 2016-07-05 16:59:41 |
Jean-Philippe Evrard |
openstack-ansible: importance |
Undecided |
High |
|
| 2016-07-06 09:21:29 |
OpenStack Infra |
openstack-ansible: status |
Confirmed |
In Progress |
|
| 2016-07-07 07:59:59 |
OpenStack Infra |
openstack-ansible: assignee |
Jean-Philippe Evrard (jean-philippe-evrard) |
Jesse Pretorius (jesse-pretorius) |
|
| 2016-07-07 11:17:09 |
OpenStack Infra |
openstack-ansible: assignee |
Jesse Pretorius (jesse-pretorius) |
Jean-Philippe Evrard (jean-philippe-evrard) |
|
| 2016-07-08 15:30:16 |
OpenStack Infra |
openstack-ansible: assignee |
Jean-Philippe Evrard (jean-philippe-evrard) |
Major Hayden (rackerhacker) |
|
| 2016-07-08 16:09:59 |
OpenStack Infra |
openstack-ansible: status |
In Progress |
Fix Released |
|
| 2017-07-12 16:41:29 |
Ihar Hrachyshka |
neutron: status |
Triaged |
Won't Fix |
|
