Bug 1281745

Summary:	Updates to HAProxy configuration
Product:	Red Hat OpenStack	Reporter:	Mike Burns <mburns>
Component:	openstack-puppet-modules	Assignee:	Emilien Macchi <emacchi>
Status:	CLOSED ERRATA	QA Contact:	Ofer Blaut <oblaut>
Severity:	unspecified	Docs Contact:
Priority:	urgent
Version:	8.0 (Liberty)	CC:	achernet, bperkins, david.costakos, hbrock, jcoufal, jschluet, mburns, mcornea, rhel-osp-director-maint, rohara, sputhenp, yeylon
Target Milestone:	ga
Target Release:	8.0 (Liberty)
Hardware:	Unspecified
OS:	Unspecified
Whiteboard:
Fixed In Version:	openstack-puppet-modules-7.0.10-1.el7ost	Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2016-04-07 21:12:10 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 Mike Burns 2015-11-13 11:25:12 UTC

Description of problem:
Based on recent field engagements, we need to update the HAProxy configuration.  The things that we need to ensure:

galera section
 stick-table type ip size should be 1000
 timeout client 60m # or higher
 timeout server 60m # or higher

on the following lines:
 server * check inter 1s port 9200 on-marked-down shutdown-sessions
need to include "backup" keyword


Version-Release number of selected component (if applicable):
7.0

How reproducible:
Always

Comment 1 chris alfonso 2015-12-07 17:17:36 UTC

Can you confirm the absence of the backup keyword renders the failover useless? If that's not the case, can you explain what the impact is?

Comment 5 Marios Andreou 2016-01-04 15:41:02 UTC

upstream bug report https://bugs.launchpad.net/tripleo/+bug/1530897 

and upstream review for the client/server timeout which afaics is the main missing thing https://review.openstack.org/#/c/263308/


The "stick-table type ip size" thing and backup are both addressed in https://review.openstack.org/#/c/194960/4/manifests/loadbalancer.pp

The only thing still missing afaics upstream is the client/server timeouts, still 0 at https://github.com/openstack/puppet-tripleo/blob/dafdc1f4efde8c57a8c7160ae0162866496089dc/manifests/loadbalancer.pp#L923

Comment 6 Marios Andreou 2016-01-25 11:34:14 UTC

the fix at https://review.openstack.org/#/c/263308 abandoned in favor of https://review.openstack.org/#/c/223223 which is merged now (and sets to 90m)

thanks

Comment 10 Ofer Blaut 2016-03-17 16:03:53 UTC

new timeouts are set on controllers


listen mysql
  bind 172.17.1.11:3306 transparent
  option tcpka
  option httpchk
  stick on dst
  stick-table type ip size 1000
  timeout client 90m
  timeout server 90m
  server overcloud-controller-0 172.17.1.13:3306 backup check fall 5 inter 2000 on-marked-down shutdown-sessions port 9200 rise 2
  server overcloud-controller-1 172.17.1.15:3306 backup check fall 5 inter 2000 on-marked-down shutdown-sessions port 9200 rise 2
  server overcloud-controller-2 172.17.1.14:3306 backup check fall 5 inter 2000 on-marked-down shutdown-sessions port 9200 rise 2


openstack-puppet-modules-7.0.13-1.el7ost.noarch

Comment 11 errata-xmlrpc 2016-04-07 21:12:10 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://rhn.redhat.com/errata/RHEA-2016-0603.html