From: John Snow <jsnow@redhat.com>
To: Sanjay Kumar <isanjaykumar@gmail.com>
Cc: qemu-devel@nongnu.org, Qemu-block <qemu-block@nongnu.org>
Subject: Re: [Qemu-devel] Query regarding qemu dirty bitmap
Date: Mon, 7 Nov 2016 16:44:50 -0500 [thread overview]
Message-ID: <cfe414b5-70b8-5159-08c2-d6395ceadee8@redhat.com> (raw)
In-Reply-To: <CAAfP9eKz2G5f0bAJ-wGA6tDcnKUEkM366Z0xDz0wPK+paQhXCw@mail.gmail.com>
Hi Sanjay, usually we reply below or in-line on this list.
On 11/05/2016 11:26 PM, Sanjay Kumar wrote:
> One more question:
>
> 4. How to interpret dirty bitmap file or get list of dirty blocks from
> the dirty bitmap in version 2.5?
>
We still don't have a working export for 2.8, what dirty bitmap file are
you referencing?
> On Sat, Nov 5, 2016 at 9:29 PM, Sanjay Kumar <isanjaykumar@gmail.com
> <mailto:isanjaykumar@gmail.com>> wrote:
>
> Hi John,
>
> Greetings. Hope you have wonderful weekend.
>
> This is Sanjay Kumar working on a backup and restore solution for KVM.
>
> The QEMU dirty bitmap QMPs support are awesome.
>
> I have below queries in regard to that.
>
> 1. How to integrate these QMPs for external backup application.
> a) Is there any command to get the list of all the dirty
> blocks?
Not yet. There are plans to extend the NBD protocol to allow a host
application to query for the status of each block. You would basically
create an NBD export of a device being tracked and then use the NBD
protocol to ask for block status.
> b) We just want list of dirty blocks, how to get those
> during migration?
Not yet possible. With qcow2 persistence you will have the ability to
see what was dirty at the time of last flush, but that is not upstream
yet. With the above NBD extension, you'll be able to see the dirty
status of any given block at any point in time you want.
>
> 2. How to create just dirty bitmap on a file? I don't need data and
> want to avoid COW.
>
You can create a dirty bitmap on any type of drive you'd like, but
persistence is only supported (in patches on-list but not yet merged)
for .qcow2, with perhaps some support for parallels in the future.
There are no plans to support bitmap persistence on raw files at present.
There have been at times some controversial discussions about allowing
qcow2 to be backed by other file formats, which would achieve what you
want -- but there is nobody actively working on this at present -- we're
more worried about getting the bitmap feature set working first.
> 3. Is there any plan to add query-block-dirty-bitmap command? If
> yes, then in which version it is coming?
>
If you mean a QMP command that will tell you which blocks are dirty, no.
We have debated fairly exhaustively and have decided that we do not want
the QMP command protocol to be used for data transfer.
We may augment existing query commands to provide some extra data in the
future, perhaps
The solution we arrived at instead was to use the NBD protocol to allow
clients to query safely the block status, but it requires a protocol
extension we're still debating on.
> Please reply. Many thanks in advance.
>
> With best regards,
> Sanjay Kumar
>
>
Good luck,
--js
prev parent reply other threads:[~2016-11-07 21:45 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-11-06 1:29 [Qemu-devel] Query regarding qemu dirty bitmap Sanjay Kumar
2016-11-06 3:26 ` Sanjay Kumar
2016-11-07 21:44 ` John Snow [this message]
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=cfe414b5-70b8-5159-08c2-d6395ceadee8@redhat.com \
--to=jsnow@redhat.com \
--cc=isanjaykumar@gmail.com \
--cc=qemu-block@nongnu.org \
--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).