Bug 1012148

Summary:	Conductor should be able to start multiple workers
Product:	[Community] RDO	Reporter:	Dan Smith <dasmith>
Component:	openstack-nova	Assignee:	Xavier Queralt <xqueralt>
Status:	CLOSED CURRENTRELEASE	QA Contact:	Ami Jeain <ajeain>
Severity:	medium	Docs Contact:
Priority:	unspecified
Version:	unspecified	CC:	apevec, dallan, ndipanov, rbryant, xqueralt, yeylon
Target Milestone:	---
Target Release:	Grizzly
Hardware:	All
OS:	Linux
Whiteboard:
Fixed In Version:	openstack-nova-2013.1.4-5	Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:	1012143	Environment:
Last Closed:	2014-03-24 07:45:12 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 Dan Smith 2013-09-25 20:13:26 UTC

+++ This bug was initially created as a clone of Bug #1012143 +++

Description of problem:

Any reasonable deployment needs more than one conductor process and grizzly did not support this natively. Right now, this is only workable if someone runs a conductor on several machines or starts multiple processes manually. We should have an /etc/sysconfig/nova-conductor config point to specify how many daemons to start so that it's easy to start multiple daemons on a single system.

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

openstack-nova-conductor-2013.1.3-2.el6

How reproducible:

Very.

Steps to Reproduce:
1. Deploy conductor on one machine
2. Deploy compute on many machines
3. Notice failures and poor performance

Actual results:

Everything funnels through a single conductor.

Expected results:

Multiple conductors should be started to spread out the load.

Additional info:

This should be pretty easy to do in the script, starting multiple workers, with log files and pid files like /var/log/nova/conductor-$i.log, etc.