[ovs-dev] [Backport Request] For branch-2.5, backport rhel-systemd integration

Russell Bryant russell at ovn.org
Sat Oct 15 01:59:30 UTC 2016

On Wed, Oct 12, 2016 at 12:03 PM, Russell Bryant <russell at ovn.org> wrote:

> On Fri, Oct 7, 2016 at 1:48 PM, Flavio Leitner <fbl at sysclose.org> wrote:
>> On Thu, Sep 08, 2016 at 07:58:21PM +0100, Markos Chandras wrote:
>> > On 09/08/2016 05:50 PM, Aaron Conole wrote:
>> > >
>> > >> It sounds like new feature territory, but you do make a case for it
>> being considered a set of fixes ...
>> > >
>> > > I agree - it straddles a line.  I was hesitant to even ask, but in
>> > > we probably need to backport these anyway, so I made an assumption
>> > > (maybe poor) that other systemd distros might use the rhel scripts and
>> > > run into this class of issues also.  Good net-izen, and all that :)
>> > >
>> > > -Aaron
>> >
>> > Hi Aaron,
>> >
>> > Thanks for the backport request. I am also interested in these fixes and
>> > it's something I could use in the SUSE package as well. For my point of
>> > view I see no blockers for those updating their 2.5.0 installations with
>> > these fixes in but I will do some testing on my end as well.
>> Most of the changes are transparent and should not cause any issues.
>> I say most of the changes because the problem is if someone is relying
>> on the openvswitch-nonetwork service state.  In that case, the service
>> doesn't exist anymore after the patchset, so it might break something.
>> I am not sure if we care about that because that service should be
>> considered internal to OVS (used only by ifcfg scripts).  The front end
>> is the openvswitch.service which remains available.
>> Regarding to be a new feature or bugfix, I think this is a bugfix for
>> two things at least.
>> 1) The real service state is represented by openvswitch service.
>> Before the patchset there were situations where the OVS threads were
>> dead but the systemd service was up & running.  The patchset intends
>> to fix this issue.
>> 2) The shutdown ordering. Before the patchset we could have OVS shutting
>> down before other networking services that depends on network.  When OVS
>> terminates, it breaks networking connectivity causing issues. The patchset
>> intends to fix the issue as well, though it needs to include the follow up
>> fix  http://openvswitch.org/pipermail/dev/2016-October/080426.html
>> In another words,
>> Acked-by: Flavio Leitner <fbl at sysclose.org>
> I'm fine with these backports.  I'll wait until we have the follow-up
> patch applied to master, which is pending your (Flavio's) Signed-off-by.

As requested, the following commits have been backported to branch-2.5.
Let me know if you run into any issues with the backports.

b23b069 ovs-ctl: Remove code for upgrading from Open vSwitch 1.9 and
bb743b6 ovs-ctl: Allow selective start for db and switch
85aa123 utilities/ovs-ctl.in: Only add_managers with vswitchd
8dfa293 utilities/ovs-ctl.in: Allow non-monitoring daemons
af3f8b0 rhel/ovsdb-server.service: Rename the nonetwork service
3704d06 rhel: Improved Systemd Integration
a3ce77d rhel-systemd: Delay shutting down the services

Russell Bryant

More information about the dev mailing list