Bug 636115

Summary:	RFE: dnsmasq: rewrite dnsmasq integration to work with a single system instance
Product:	[Community] Virtualization Tools	Reporter:	Daniel Berrangé <berrange>
Component:	libvirt	Assignee:	Libvirt Maintainers <libvirt-maint>
Status:	CLOSED DEFERRED	QA Contact:
Severity:	medium	Docs Contact:
Priority:	low
Version:	unspecified	CC:	crobinso, libvirt-maint, serge, xen-maint
Target Milestone:	---
Target Release:	---
Hardware:	All
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2016-05-02 14:18:08 UTC	Type:	---
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:
Bug Depends On:
Bug Blocks:	636029

Description Daniel Berrangé 2010-09-21 14:28:53 UTC

Description of problem:
Upstream dnsmasq offered to extend its functionality to allow libvirt to drop in configuration for a single system instance. This will avoid the problem of clashing dnsmsaq processes on a single host.

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


How reproducible:


Steps to Reproduce:
1.
2.
3.
  
Actual results:


Expected results:


Additional info:

Comment 1 serge 2015-08-27 17:00:21 UTC

Hi,

commit http://libvirt.org/git/?p=libvirt.git;a=commit;h=df3de82c06d42fed832c1955dafdee85879bd15a

changed behavior to always start dnsmasq if ip addresses are defined.  There is no keyword (that I can find) to tell libvirt "don't start a dnsmasq for this bridge, I want to run one myself".  Is adding such a keyword ok, or is it nacked on principle?

Comment 2 Cole Robinson 2016-03-20 23:13:08 UTC

(In reply to serge from comment #1)
> Hi,
> 
> commit
> http://libvirt.org/git/?p=libvirt.git;a=commit;
> h=df3de82c06d42fed832c1955dafdee85879bd15a
> 
> changed behavior to always start dnsmasq if ip addresses are defined.  There
> is no keyword (that I can find) to tell libvirt "don't start a dnsmasq for
> this bridge, I want to run one myself".  Is adding such a keyword ok, or is
> it nacked on principle?

serge if your issue was never resolved I suggest starting a thread on libvir-list instead, it will probably get a better response

Comment 3 Cole Robinson 2016-03-20 23:14:40 UTC

I know we've reworked dnsmasq code quite a bit over the years, but we still haven't gotten around to using a single system dnsmasq instance and I don't really hear people clamoring for it... is this even still relevant?

Comment 4 Cole Robinson 2016-05-02 14:18:08 UTC

Just re-iterating that I don't really see us reworking the network driver to work like this. Maybe we could make it an optional config to use a single system daemon, but since it doesn't sound like there's many requests for this feature and the work isn't exactly forthcoming I don't see the benefit in tracking it here