qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: "Daniel P. Berrangé" <berrange@redhat.com>
To: Fabiano Rosas <farosas@suse.de>
Cc: qemu-devel@nongnu.org, armbru@redhat.com,
	Juan Quintela <quintela@redhat.com>, Peter Xu <peterx@redhat.com>,
	Leonardo Bras <leobras@redhat.com>,
	Claudio Fontana <cfontana@suse.de>,
	Eric Blake <eblake@redhat.com>
Subject: Re: [PATCH v2 28/29] migration: Add direct-io parameter
Date: Wed, 25 Oct 2023 16:22:05 +0100	[thread overview]
Message-ID: <ZTkynR7UelpQrpqA@redhat.com> (raw)
In-Reply-To: <87a5s6ep1z.fsf@suse.de>

On Wed, Oct 25, 2023 at 11:48:08AM -0300, Fabiano Rosas wrote:
> Daniel P. Berrangé <berrange@redhat.com> writes:
> 
> > On Mon, Oct 23, 2023 at 05:36:07PM -0300, Fabiano Rosas wrote:
> >> Add the direct-io migration parameter that tells the migration code to
> >> use O_DIRECT when opening the migration stream file whenever possible.
> >> 
> >> This is currently only used for the secondary channels of fixed-ram
> >> migration, which can guarantee that writes are page aligned.
> >
> > When you say "secondary channels", I presume you're meaning that
> > the bulk memory regions will be written with O_DIRECT, while
> > general vmstate will use normal I/O on the main channel ?  If so,
> > could we explain that a little more directly.
> 
> Yes, the main channel writes via QEMUFile, so no O_DIRECT. The channels
> created via multifd_new_send_channel_create() have O_DIRECT enabled.
> 
> > Having a mixture of O_DIRECT and non-O_DIRECT I/O on the same
> > file is a little bit of an unusual situation. It will work for
> > us because we're writing to different regions of the file in
> > each case.
> >
> > Still I wonder if it would be sane in the outgoing case to
> > include a fsync() on the file in the main channel, to guarantee
> > that the whole saved file is on-media at completion ? Or perhaps
> > suggest in QAPI that mgmts might consider doing a fsync
> > themselves ?
> 
> I think that should be present in QIOChannelFile in general. Not even
> related to this series. I'll add it at qio_channel_file_close() unless
> you object.

Yes, looking at the places QIOChannelFile is used, I think they would
all benefit from fdatasync().  fsync() is probably too big of a hammer

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:[~2023-10-25 15:22 UTC|newest]

Thread overview: 128+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-10-23 20:35 [PATCH v2 00/29] migration: File based migration with multifd and fixed-ram Fabiano Rosas
2023-10-23 20:35 ` [PATCH v2 01/29] tests/qtest: migration events Fabiano Rosas
2023-10-25  9:44   ` Thomas Huth
2023-10-25 10:14   ` Daniel P. Berrangé
2023-10-25 13:21   ` Fabiano Rosas
2023-10-25 13:33     ` Steven Sistare
2023-10-23 20:35 ` [PATCH v2 02/29] tests/qtest: Move QTestMigrationState to libqtest Fabiano Rosas
2023-10-25 10:17   ` Daniel P. Berrangé
2023-10-25 13:19     ` Fabiano Rosas
2023-10-23 20:35 ` [PATCH v2 03/29] tests/qtest: Allow waiting for migration events Fabiano Rosas
2023-10-23 20:35 ` [PATCH v2 04/29] migration: Return the saved state from global_state_store Fabiano Rosas
2023-10-25 10:19   ` Daniel P. Berrangé
2023-10-23 20:35 ` [PATCH v2 05/29] migration: Introduce global_state_store_once Fabiano Rosas
2023-10-23 20:35 ` [PATCH v2 06/29] migration: Add auto-pause capability Fabiano Rosas
2023-10-24  5:25   ` Markus Armbruster
2023-10-24 18:12     ` Fabiano Rosas
2023-10-25  5:33       ` Markus Armbruster
2023-10-25  8:48   ` Daniel P. Berrangé
2023-10-25 13:57     ` Fabiano Rosas
2023-10-25 14:20       ` Daniel P. Berrangé
2023-10-25 14:58         ` Peter Xu
2023-10-25 15:25           ` Daniel P. Berrangé
2023-10-25 15:36             ` Peter Xu
2023-10-25 15:40               ` Daniel P. Berrangé
2023-10-25 17:20                 ` Peter Xu
2023-10-25 17:31                   ` Daniel P. Berrangé
2023-10-25 19:28                     ` Peter Xu
2023-10-23 20:35 ` [PATCH v2 07/29] migration: Run "file:" migration with a stopped VM Fabiano Rosas
2023-10-23 20:35 ` [PATCH v2 08/29] tests/qtest: File migration auto-pause tests Fabiano Rosas
2023-10-23 20:35 ` [PATCH v2 09/29] io: add and implement QIO_CHANNEL_FEATURE_SEEKABLE for channel file Fabiano Rosas
2023-10-23 20:35 ` [PATCH v2 10/29] io: Add generic pwritev/preadv interface Fabiano Rosas
2023-10-24  8:18   ` Daniel P. Berrangé
2023-10-24 19:06     ` Fabiano Rosas
2023-10-23 20:35 ` [PATCH v2 11/29] io: implement io_pwritev/preadv for QIOChannelFile Fabiano Rosas
2023-10-23 20:35 ` [PATCH v2 12/29] migration/qemu-file: add utility methods for working with seekable channels Fabiano Rosas
2023-10-25 10:22   ` Daniel P. Berrangé
2023-10-23 20:35 ` [PATCH v2 13/29] migration: fixed-ram: Add URI compatibility check Fabiano Rosas
2023-10-25 10:27   ` Daniel P. Berrangé
2023-10-31 16:06   ` Peter Xu
2023-10-23 20:35 ` [PATCH v2 14/29] migration/ram: Introduce 'fixed-ram' migration capability Fabiano Rosas
2023-10-24  5:33   ` Markus Armbruster
2023-10-24 18:35     ` Fabiano Rosas
2023-10-25  6:18       ` Markus Armbruster
2023-10-23 20:35 ` [PATCH v2 15/29] migration/ram: Add support for 'fixed-ram' outgoing migration Fabiano Rosas
2023-10-25  9:39   ` Daniel P. Berrangé
2023-10-25 14:03     ` Fabiano Rosas
2023-11-01 15:23     ` Peter Xu
2023-11-01 15:52       ` Daniel P. Berrangé
2023-11-01 16:24         ` Peter Xu
2023-11-01 16:37           ` Daniel P. Berrangé
2023-11-01 17:30             ` Peter Xu
2023-10-31 16:52   ` Peter Xu
2023-10-31 17:33     ` Fabiano Rosas
2023-10-31 17:59       ` Peter Xu
2023-10-23 20:35 ` [PATCH v2 16/29] migration/ram: Add support for 'fixed-ram' migration restore Fabiano Rosas
2023-10-25  9:43   ` Daniel P. Berrangé
2023-10-25 14:07     ` Fabiano Rosas
2023-10-31 19:03       ` Peter Xu
2023-11-01  9:26         ` Daniel P. Berrangé
2023-11-01 14:21           ` Peter Xu
2023-11-01 14:28             ` Daniel P. Berrangé
2023-11-01 15:00               ` Peter Xu
2023-11-06 13:18                 ` Fabiano Rosas
2023-11-06 21:00                   ` Peter Xu
2023-11-07  9:02                     ` Daniel P. Berrangé
2023-10-31 19:09   ` Peter Xu
2023-10-31 20:00     ` Fabiano Rosas
2023-10-23 20:35 ` [PATCH v2 17/29] tests/qtest: migration-test: Add tests for fixed-ram file-based migration Fabiano Rosas
2023-10-23 20:35 ` [PATCH v2 18/29] migration/multifd: Allow multifd without packets Fabiano Rosas
2023-10-23 20:35 ` [PATCH v2 19/29] migration/multifd: Add outgoing QIOChannelFile support Fabiano Rosas
2023-10-25  9:52   ` Daniel P. Berrangé
2023-10-25 14:12     ` Fabiano Rosas
2023-10-25 14:23       ` Daniel P. Berrangé
2023-10-25 15:00         ` Fabiano Rosas
2023-10-25 15:26           ` Daniel P. Berrangé
2023-10-31 20:11   ` Peter Xu
2023-10-23 20:35 ` [PATCH v2 20/29] migration/multifd: Add incoming " Fabiano Rosas
2023-10-25 10:29   ` Daniel P. Berrangé
2023-10-25 14:18     ` Fabiano Rosas
2023-10-31 21:28   ` Peter Xu
2023-10-23 20:36 ` [PATCH v2 21/29] migration/multifd: Add pages to the receiving side Fabiano Rosas
2023-10-31 22:10   ` Peter Xu
2023-10-31 23:18     ` Fabiano Rosas
2023-11-01 15:55       ` Peter Xu
2023-11-01 17:20         ` Fabiano Rosas
2023-11-01 17:35           ` Peter Xu
2023-11-01 18:14             ` Fabiano Rosas
2023-10-23 20:36 ` [PATCH v2 22/29] io: Add a pwritev/preadv version that takes a discontiguous iovec Fabiano Rosas
2023-10-24  8:50   ` Daniel P. Berrangé
2023-10-23 20:36 ` [PATCH v2 23/29] migration/ram: Add a wrapper for fixed-ram shadow bitmap Fabiano Rosas
2023-11-01 14:29   ` Peter Xu
2023-10-23 20:36 ` [PATCH v2 24/29] migration/ram: Ignore multifd flush when doing fixed-ram migration Fabiano Rosas
2023-10-25  9:09   ` Daniel P. Berrangé
2023-10-23 20:36 ` [PATCH v2 25/29] migration/multifd: Support outgoing fixed-ram stream format Fabiano Rosas
2023-10-25  9:23   ` Daniel P. Berrangé
2023-10-25 14:21     ` Fabiano Rosas
2023-10-23 20:36 ` [PATCH v2 26/29] migration/multifd: Support incoming " Fabiano Rosas
2023-10-23 20:36 ` [PATCH v2 27/29] tests/qtest: Add a multifd + fixed-ram migration test Fabiano Rosas
2023-10-23 20:36 ` [PATCH v2 28/29] migration: Add direct-io parameter Fabiano Rosas
2023-10-24  5:41   ` Markus Armbruster
2023-10-24 19:32     ` Fabiano Rosas
2023-10-25  6:23       ` Markus Armbruster
2023-10-25  8:44       ` Daniel P. Berrangé
2023-10-25 14:32         ` Fabiano Rosas
2023-10-25 14:43           ` Daniel P. Berrangé
2023-10-25 17:30             ` Fabiano Rosas
2023-10-25 17:45               ` Daniel P. Berrangé
2023-10-25 18:10                 ` Fabiano Rosas
2023-10-30 22:51             ` Fabiano Rosas
2023-10-31  9:03               ` Daniel P. Berrangé
2023-10-31 13:05                 ` Fabiano Rosas
2023-10-31 13:45                   ` Daniel P. Berrangé
2023-10-31 14:33                     ` Fabiano Rosas
2023-10-31 15:22                       ` Daniel P. Berrangé
2023-10-31 15:52                         ` Fabiano Rosas
2023-10-31 15:58                           ` Daniel P. Berrangé
2023-10-31 19:05                             ` Fabiano Rosas
2023-11-01  9:30                               ` Daniel P. Berrangé
2023-11-01 12:16                                 ` Fabiano Rosas
2023-11-01 12:23                                   ` Daniel P. Berrangé
2023-11-01 12:30                                     ` Fabiano Rosas
2023-10-24  8:33   ` Daniel P. Berrangé
2023-10-24 19:06     ` Fabiano Rosas
2023-10-25  9:07   ` Daniel P. Berrangé
2023-10-25 14:48     ` Fabiano Rosas
2023-10-25 15:22       ` Daniel P. Berrangé [this message]
2023-10-23 20:36 ` [PATCH v2 29/29] tests/qtest: Add a test for migration with direct-io and multifd Fabiano Rosas
2023-10-25  9:25   ` Daniel P. Berrangé

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=ZTkynR7UelpQrpqA@redhat.com \
    --to=berrange@redhat.com \
    --cc=armbru@redhat.com \
    --cc=cfontana@suse.de \
    --cc=eblake@redhat.com \
    --cc=farosas@suse.de \
    --cc=leobras@redhat.com \
    --cc=peterx@redhat.com \
    --cc=qemu-devel@nongnu.org \
    --cc=quintela@redhat.com \
    /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).