From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-0.6 required=3.0 tests=DKIM_INVALID,DKIM_SIGNED, HEADER_FROM_DIFFERENT_DOMAINS,MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS autolearn=no autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 52959C433DF for ; Fri, 5 Jun 2020 10:08:32 +0000 (UTC) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 188A62053B for ; Fri, 5 Jun 2020 10:08:32 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=fail reason="signature verification failed" (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b="GHhqoc+T" DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 188A62053B Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=redhat.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Received: from localhost ([::1]:45826 helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1jh9H5-0006Th-De for qemu-devel@archiver.kernel.org; Fri, 05 Jun 2020 06:08:31 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]:43610) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1jh9GW-0005nz-1n for qemu-devel@nongnu.org; Fri, 05 Jun 2020 06:07:56 -0400 Received: from us-smtp-delivery-1.mimecast.com ([205.139.110.120]:24237 helo=us-smtp-1.mimecast.com) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_CBC_SHA1:256) (Exim 4.90_1) (envelope-from ) id 1jh9GV-0003Fn-2i for qemu-devel@nongnu.org; Fri, 05 Jun 2020 06:07:55 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1591351674; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: in-reply-to:in-reply-to:references:references; bh=jTzDjQ320wWxNkVdVwZYN6o/l9ZaEQBGHF5094Y7wNo=; b=GHhqoc+T7Sc3qm2kUM1WItKv9fFW6gpknpcYZ6DrxIUS3Jcm7Ib3NRkxEku7IjYmcwONLp pfoUr2z4xe5Gp124FYYq79EVj+21y4qiI6BO+rZZokRVk2RZkFBwtp5uKNIQlAsPhfsdJK gmlqRfGzX0uh1yI2LedvzH2UT+9HPjE= Received: from mimecast-mx01.redhat.com (mimecast-mx01.redhat.com [209.132.183.4]) (Using TLS) by relay.mimecast.com with ESMTP id us-mta-158-NmJHcWAQPj-kLcu_Gp5oUw-1; Fri, 05 Jun 2020 06:07:52 -0400 X-MC-Unique: NmJHcWAQPj-kLcu_Gp5oUw-1 Received: from smtp.corp.redhat.com (int-mx08.intmail.prod.int.phx2.redhat.com [10.5.11.23]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx01.redhat.com (Postfix) with ESMTPS id F2FFB1854C5A; Fri, 5 Jun 2020 10:07:50 +0000 (UTC) Received: from linux.fritz.box (ovpn-113-112.ams2.redhat.com [10.36.113.112]) by smtp.corp.redhat.com (Postfix) with ESMTPS id 0BAA0196F7; Fri, 5 Jun 2020 10:07:48 +0000 (UTC) Date: Fri, 5 Jun 2020 12:07:47 +0200 From: Kevin Wolf To: Peter Krempa Subject: Re: [PATCH RFC v2 1/5] block: add bitmap-populate job Message-ID: <20200605100747.GH5869@linux.fritz.box> References: <20200514034922.24834-2-jsnow@redhat.com> <20200604091231.GC4512@linux.fritz.box> <20200604091651.GF2995787@angien.pipo.sk> <20200604113145.GE4512@linux.fritz.box> <20200604162237.GA22354@angien.pipo.sk> <20200605090123.GD5869@linux.fritz.box> <20200605092403.GB22354@angien.pipo.sk> <20200605094407.GG5869@linux.fritz.box> <20200605095810.GC22354@angien.pipo.sk> MIME-Version: 1.0 In-Reply-To: <20200605095810.GC22354@angien.pipo.sk> X-Scanned-By: MIMEDefang 2.84 on 10.5.11.23 X-Mimecast-Spam-Score: 0 X-Mimecast-Originator: redhat.com Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Received-SPF: pass client-ip=205.139.110.120; envelope-from=kwolf@redhat.com; helo=us-smtp-1.mimecast.com X-detected-operating-system: by eggs.gnu.org: First seen = 2020/06/05 03:27:26 X-ACL-Warn: Detected OS = Linux 2.2.x-3.x [generic] [fuzzy] X-Spam_score_int: -20 X-Spam_score: -2.1 X-Spam_bar: -- X-Spam_report: (-2.1 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=0.001, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_NONE=-0.0001, RCVD_IN_MSPIKE_H4=0.001, RCVD_IN_MSPIKE_WL=0.001, SPF_PASS=-0.001 autolearn=_AUTOLEARN X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: vsementsov@virtuozzo.com, Eduardo Habkost , qemu-block@nongnu.org, qemu-devel@nongnu.org, Markus Armbruster , Cleber Rosa , Max Reitz , John Snow Errors-To: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Sender: "Qemu-devel" Am 05.06.2020 um 11:58 hat Peter Krempa geschrieben: > On Fri, Jun 05, 2020 at 11:44:07 +0200, Kevin Wolf wrote: > > Am 05.06.2020 um 11:24 hat Peter Krempa geschrieben: > > > On Fri, Jun 05, 2020 at 11:01:23 +0200, Kevin Wolf wrote: > > > > Am 04.06.2020 um 18:22 hat Peter Krempa geschrieben: > > > > > On Thu, Jun 04, 2020 at 13:31:45 +0200, Kevin Wolf wrote: > > > > > > Am 04.06.2020 um 11:16 hat Peter Krempa geschrieben: > > > > > > > On Thu, Jun 04, 2020 at 11:12:31 +0200, Kevin Wolf wrote: > > > > > > > > Am 18.05.2020 um 22:49 hat Eric Blake geschrieben: > > > > > > > > > > + > > > > > > > > > > + /* NB: new bitmap is anonymous and enabled */ > > > > > > > > > > + cluster_size = bdrv_dirty_bitmap_granularity(target_bitmap); > > > > > > > > > > + new_bitmap = bdrv_create_dirty_bitmap(bs, cluster_size, NULL, errp); > > > > > > > > > > + if (!new_bitmap) { > > > > > > > > > > + return NULL; > > > > > > > > > > + } > > > > > > > > > > > > > > > > > > This means if the guest writes to the disk while the job is ongoing, the > > > > > > > > > bitmap will be updated to mark that portion of the bitmap as set, even if it > > > > > > > > > was not allocated at the time the job started. But then again, the guest > > > > > > > > > writes are causing allocation, so this seems like the right thing to do. > > > > > > > > > > > > > > > > Is the target bitmap active at the same time, i.e. will it get the > > > > > > > > correct information only from new_bitmap or are the bits already set in > > > > > > > > it anyway? > > > > > > > > > > > > > > Yes, libvirt plans to use it with an active non-persistent bitmap which > > > > > > > will in subsequent steps be merged into others. The bitmap is added in > > > > > > > the same transaction. The bitmap must be active, because we need to wait > > > > > > > for the block jobs to finish before it becomes usable and thus can't > > > > > > > sequence in other operations until later. > > > > > > > > > > > > A lot of bitmap merging then, because the block job in this series > > > > > > already creates a temporary internal bitmap that is merged into the > > > > > > target bitmap on completion. But if the target bitmap is only libvirt's > > > > > > temporary bitmap to be merged to yet another bitmap, I wonder if this > > > > > > process shouldn't be simplified. > > > > > > > > > > Possibly yes, but I'll leave that for later. All of this is done when > > > > > executin very expensive operations anyways so for our first > > > > > implementation it IMO won't matter that much. > > > > > > > > I'm not necessarily saying that the change is needed on the libvirt > > > > side. It could also be that the block job should directly work with the > > > > given bitmap instead of having its internal temporary bitmap. Changing > > > > this later would mean changing the semantics of the block job, so it > > > > would be somewhat problematic. > > > > > > > > It would be good to have a clear picture of what we want the final > > > > result to look like. > > > > > > Well with current semantics of the 'nodename' argument controling both > > > where the populated bitmap is located and also which node's allocation > > > bitmap to take I don't think we can optimize it further in libvirt. > > > > > > Current usage scenario is that we use a temporary bitmap populated with > > > the job to merge with bitmaps present in nodes which are removed by > > > blockjobs into the destination node of the block job. This means that > > > the real destination of the bits populated is in a different node than > > > it was originally and the above job semantics don't allow that. > > > > So does this mean that a better API wouldn't only take a node-name and > > bitmap name (where the node identified by node-name is not only where > > the target bitmap is, but also the node whose allocation status is > > queried), but that it should take two different node-names for source > > (= reading allocation status) and target (= owner of the bitmap)? > > Yes. That way one of the merges would be merged (heh) into the operation > itself preventing us from the need to have an extra temporary bitmap. > > > > Either way I'd strongly prefer to be able to kick off all the populate > > > jobs at once rather than having to sequence them so any semantic change > > > towards making it possible to target bitmaps in a different node would > > > also require that multiple jobs can run in parallel with a single bitmap > > > as destination. I'm not sure if that doesn't overcomplicate things > > > though. > > > > Other people are more familiar with the dirty bitmap code, so I may be > > wrong, but intuitively, I don't see any problem with multiple jobs > > dirtying blocks in the same bitmap. Or, with the internal temporary > > bitmap as used in this version of the series, multiple jobs that, one > > after another, merge their result to the same bitmap on completion. > > I don't see a problem with the bitmaps itself since we are only OR-ing > results together. I just wanted to state the desired usage. > > The above was actually inspired by a very recent problem I have in my > attempt to use the dirty bitmap populate job to refactor how libvirt > handles bitmaps. I've just figured out that I need to shuffle around > some stuff as I can't run the dirty-bitmap-populate job while an active > layer commit is in synchronised phase and I wanted to do the merging at > that point. That reminded me of a possible gotcha in having to sequence > the blockjobs which certainly would be more painful. It would probably be good to have not only an iotests case that tests the low-level functionality of the block job, but also one that resembles the way libvirt would actually use it in combination with other jobs. Kevin