On Tue, Jan 24, 2023, 04:50 Dr. David Alan Gilbert <dgilb...@redhat.com> wrote:
> * Stefan Hajnoczi (stefa...@gmail.com) wrote: > > On Mon, 23 Jan 2023 at 14:54, Stefan Hajnoczi <stefa...@redhat.com> > wrote: > > > > > > On Mon, Jan 23, 2023 at 06:27:23PM +0000, Dr. David Alan Gilbert wrote: > > > > * Michael S. Tsirkin (m...@redhat.com) wrote: > > > > > On Sun, Jan 22, 2023 at 06:09:40PM +0200, Anton Kuchin wrote: > > > > > > > > > > > > On 22/01/2023 16:46, Michael S. Tsirkin wrote: > > > > > > > On Sun, Jan 22, 2023 at 02:36:04PM +0200, Anton Kuchin wrote: > > > > > > > > > > This flag should be set when qemu don't need to worry > about any > > > > > > > > > > external state stored in vhost-user daemons during > migration: > > > > > > > > > > don't fail migration, just pack generic virtio device > states to > > > > > > > > > > migration stream and orchestrator guarantees that the > rest of the > > > > > > > > > > state will be present at the destination to restore full > context and > > > > > > > > > > continue running. > > > > > > > > > Sorry I still do not get it. So fundamentally, why do we > need this property? > > > > > > > > > vhost-user-fs is not created by default that we'd then > need opt-in to > > > > > > > > > the special "migrateable" case. > > > > > > > > > That's why I said it might make some sense as a device > property as qemu > > > > > > > > > tracks whether device is unplugged for us. > > > > > > > > > > > > > > > > > > But as written, if you are going to teach the orchestrator > about > > > > > > > > > vhost-user-fs and its special needs, just teach it when to > migrate and > > > > > > > > > where not to migrate. > > > > > > > > > > > > > > > > > > Either we describe the special situation to qemu and let > qemu > > > > > > > > > make an intelligent decision whether to allow migration, > > > > > > > > > or we trust the orchestrator. And if it's the latter, then > 'migrate' > > > > > > > > > already says orchestrator decided to migrate. > > > > > > > > The problem I'm trying to solve is that most of vhost-user > devices > > > > > > > > now block migration in qemu. And this makes sense since qemu > can't > > > > > > > > extract and transfer backend daemon state. But this prevents > us from > > > > > > > > updating qemu executable via local migration. So this flag is > > > > > > > > intended more as a safety check that says "I know what I'm > doing". > > > > > > > > > > > > > > > > I agree that it is not really necessary if we trust the > orchestrator > > > > > > > > to request migration only when the migration can be > performed in a > > > > > > > > safe way. But changing the current behavior of vhost-user-fs > from > > > > > > > > "always blocks migration" to "migrates partial state whenever > > > > > > > > orchestrator requests it" seems a little dangerous and can > be > > > > > > > > misinterpreted as full support for migration in all cases. > > > > > > > It's not really different from block is it? orchestrator has > to arrange > > > > > > > for backend migration. I think we just assumed there's no > use-case where > > > > > > > this is practical for vhost-user-fs so we blocked it. > > > > > > > But in any case it's orchestrator's responsibility. > > > > > > > > > > > > Yes, you are right. So do you think we should just drop the > blocker > > > > > > without adding a new flag? > > > > > > > > > > I'd be inclined to. I am curious what do dgilbert and stefanha > think though. > > > > > > > > Yes I think that's probably OK, as long as we use the flag for > knowing > > > > how to handle the discard bitmap as a proxy for the daemon knowing > how > > > > to handle *some* migrations; knowing which migrations is then the job > > > > for the orchestrator to be careful of. > > > > > > I think the feature bit is not a good way to detect live migration > > > support. vhost-user backends typically use libvhost-user, rust-vmm's > > > vhost-user-backend crate, etc where this feature can be implemented for > > > free. If the feature bit is advertized we don't know if the device > > > implementation (net, blk, fs, etc) is aware of migration at all. > > > > I checked how bad the situation is. libvhost-user currently enables > > LOG_ALL by default. :( > > > > So I don't think the front-end can use LOG_ALL alone to determine > > whether or not migration is supported by the back-end. > > > > There are several existing back-ends based on libvhost-user that have > > no concept of reconnection or migration but report the LOG_ALL feature > > bit. > > Ouch, yes that's messy. > > Going back to the original question; I don't think a command line flag > will work though, because even for a given VM there's the possibility > of some (local) migrations working but other (remote) migrations not > working; so you don't know at the point you start the VM whether > your migrations are going to work. > The user or management tool should know which types of migration a vhost-user-fs backend supports. That can be passed in as a per-device parameter. Then a migration parameter can be used to distinguish between same host and remote host migration? QEMU already distinguishes between pre-copy and post-copy migration, so this can be thought of as yet another type of migration. Stefan >
