qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: "Daniel P. Berrangé" <berrange@redhat.com>
To: "Denis V. Lunev" <den@virtuozzo.com>
Cc: Efim Shevrin <efim.shevrin@virtuozzo.com>,
	qemu-devel@nongnu.org, kkostiuk@redhat.com, michael.roth@amd.com,
	pbonzini@redhat.com, den@openvz.org
Subject: Re: [PATCH] qga: Start qemu-ga service after NetworkManager start
Date: Thu, 17 Aug 2023 12:24:58 +0100	[thread overview]
Message-ID: <ZN4DipkE0wPn5H0G@redhat.com> (raw)
In-Reply-To: <59a42366-58e2-7147-1bd4-e85b446b0477@virtuozzo.com>

On Thu, Aug 17, 2023 at 12:47:48PM +0200, Denis V. Lunev wrote:
> On 8/17/23 12:26, Daniel P. Berrangé wrote:
> > On Thu, Aug 17, 2023 at 04:04:46AM +0800, Efim Shevrin via wrote:
> > > From: Fima Shevrin <efim.shevrin@virtuozzo.com>
> > > 
> > > When the guest OS starts, qemu-ga sends an event to the host.
> > > This event allows services on the host to start configuring
> > > the already running guest OS. When configuring network settings,
> > > it is possible that an external service will receive a signal
> > > from qemu-ga about the start of guest OS, while NetworkManager
> > > may not be running yet. Therefore, network setting may not
> > > be available. With the current patch, we eliminate the described
> > > race condition between qemu-ga and NetworkManager for guest OS
> > > network setting cases.
> > A host mgmt app should not be assuming that networking is
> > running at all, and if it is, the guest might not even be
> > on the same subnet as the host. Conceivably someone could
> > even be using QGA to inject files to configure networking.
> > 
> > Overall, IMHO, it is wrong to correlate host notification
> > that QGA is running, with any assumption about state of
> > the guest networking.
> > 
> > IOW, I don't think we should make an attempt to serialize
> > startup of QGA after guest networking. The QGA should be
> > run as soon as its data transport (virtio-serial/vsock)
> > is available, which is the current state.
> > 
> > If a host consumer of QGA wants to do something with
> > guest networking, they should check for guest network
> > state with the 'guest-network-get-interfaces' command.
> 
> This is very tight question at my opinion.
> 
> Frankly speaking my original intention here would be
> to put QGA at the very end of the startup sequence
> in order to avoid complications. In the other case
> any host side software, which would like to control
> the guest behavior will be in trouble.
> 
> For example, yet another common application is a
> guest backup, here I would say application based
> backup. In a normal case this is done through
>    /etc/qemu-ga/fsfreeze-hook

It will already be started after local-fs.target, since it does not
have DefaultDependencies=no set. This should be sufficient for
fsfreeze use cases IIUC.

> I would say that agent should start very last, in the
> other case *ANY* software working through the agent
> must take efforts to avoid such transient state, which
> is complicated.



With regards,
Daniel
-- 
|: https://berrange.com      -o-    https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org         -o-            https://fstop138.berrange.com :|
|: https://entangle-photo.org    -o-    https://www.instagram.com/dberrange :|



  reply	other threads:[~2023-08-17 11:26 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-08-16 20:04 [PATCH] qga: Start qemu-ga service after NetworkManager start Efim Shevrin via
2023-08-17  6:15 ` Konstantin Kostiuk
2023-08-17 10:28   ` Daniel P. Berrangé
2023-08-17 10:26 ` Daniel P. Berrangé
2023-08-17 10:47   ` Denis V. Lunev
2023-08-17 11:24     ` Daniel P. Berrangé [this message]
2023-08-17 10:53   ` Michael Tokarev

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=ZN4DipkE0wPn5H0G@redhat.com \
    --to=berrange@redhat.com \
    --cc=den@openvz.org \
    --cc=den@virtuozzo.com \
    --cc=efim.shevrin@virtuozzo.com \
    --cc=kkostiuk@redhat.com \
    --cc=michael.roth@amd.com \
    --cc=pbonzini@redhat.com \
    --cc=qemu-devel@nongnu.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).