[prev in list] [next in list] [prev in thread] [next in thread] 

List:       linux-virtualization
Subject:    Re: [summary] virtio network device failover writeup
From:       Liran Alon <liran.alon () oracle ! com>
Date:       2019-03-21 17:15:19
Message-ID: 65C9DA4E-D959-4F9D-AD5B-1831387E652E () oracle ! com
[Download RAW message or body]



> On 21 Mar 2019, at 19:12, Michael S. Tsirkin <mst@redhat.com> wrote:
> 
> On Thu, Mar 21, 2019 at 06:31:35PM +0200, Liran Alon wrote:
> > 
> > 
> > > On 21 Mar 2019, at 17:50, Michael S. Tsirkin <mst@redhat.com> wrote:
> > > 
> > > On Thu, Mar 21, 2019 at 08:45:17AM -0700, Stephen Hemminger wrote:
> > > > On Thu, 21 Mar 2019 15:04:37 +0200
> > > > Liran Alon <liran.alon@oracle.com> wrote:
> > > > 
> > > > > > 
> > > > > > OK. Now what happens if master is moved to another namespace? Do we need
> > > > > > to move the slaves too?  
> > > > > 
> > > > > No. Why would we move the slaves? The whole point is to make most customer \
> > > > > ignore the net-failover slaves and remain them "hidden" in their dedicated \
> > > > > netns. We won't prevent customer from explicitly moving the net-failover \
> > > > > slaves out of this netns, but we will not move them out of there \
> > > > > automatically.
> > > > 
> > > > 
> > > > The 2-device netvsc already handles case where master changes namespace.
> > > 
> > > Is it by moving slave with it?
> > 
> > See c0a41b887ce6 ("hv_netvsc: move VF to same namespace as netvsc device").
> > It seems that when NetVSC master netdev changes netns, the VF is moved to the \
> > same netns by the NetVSC driver. Kinda the opposite than what we are suggesting \
> > here to make sure that the net-failover master netdev is on a separate netns than \
> > it's slaves... 
> > -Liran
> > 
> > > 
> > > -- 
> > > MST
> 
> Not exactly opposite I'd say.
> 
> If failover is in host ns, slaves in /primary and /standby, then moving
> failover to /container should move slaves to /container/primary and
> /container/standby.

Yes I agree.
I meant that they tried to keep the VF on the same netns as the NetVSC.
But of course what you just described is exactly the functionality I would have \
wanted in our net-failover mechanism.

-Liran

> 
> 
> -- 
> MST

_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization


[prev in list] [next in list] [prev in thread] [next in thread]