From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:41059) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1Zvk0K-0000uB-61 for qemu-devel@nongnu.org; Mon, 09 Nov 2015 05:48:53 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1Zvk0G-0007av-23 for qemu-devel@nongnu.org; Mon, 09 Nov 2015 05:48:52 -0500 Received: from mailpro.odiso.net ([89.248.209.98]:40682) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1Zvk0F-0007ap-Rw for qemu-devel@nongnu.org; Mon, 09 Nov 2015 05:48:47 -0500 Date: Mon, 9 Nov 2015 11:48:45 +0100 (CET) From: Alexandre DERUMIER Message-ID: <253899541.6094258.1447066125770.JavaMail.zimbra@oxygem.tv> In-Reply-To: <564049BA.1090805@parallels.com> References: <310812067.2785504.1447038645014.JavaMail.zimbra@oxygem.tv> <5640490B.5030700@parallels.com> <564049BA.1090805@parallels.com> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable Subject: Re: [Qemu-devel] qemu : rbd block driver internal snapshot and vm_stop is hanging forever List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: ceph-devel Cc: jdurgin@redhat.com, qemu-devel adding to ceph.conf [client] rbd_non_blocking_aio =3D false fix the problem for me (with rbd_cache=3Dfalse) (@cc jdurgin@redhat.com) ----- Mail original ----- De: "Denis V. Lunev" =C3=80: "aderumier" , "ceph-devel" , "qemu-devel" Envoy=C3=A9: Lundi 9 Novembre 2015 08:22:34 Objet: Re: [Qemu-devel] qemu : rbd block driver internal snapshot and vm_st= op is hanging forever On 11/09/2015 10:19 AM, Denis V. Lunev wrote:=20 > On 11/09/2015 06:10 AM, Alexandre DERUMIER wrote:=20 >> Hi,=20 >>=20 >> with qemu (2.4.1), if I do an internal snapshot of an rbd device,=20 >> then I pause the vm with vm_stop,=20 >>=20 >> the qemu process is hanging forever=20 >>=20 >>=20 >> monitor commands to reproduce:=20 >>=20 >>=20 >> # snapshot_blkdev_internal drive-virtio0 yoursnapname=20 >> # stop=20 >>=20 >>=20 >>=20 >>=20 >> I don't see this with qcow2 or sheepdog block driver for example.=20 >>=20 >>=20 >> Regards,=20 >>=20 >> Alexandre=20 >>=20 > this could look like the problem I have recenty trying to=20 > fix with dataplane enabled. Patch series is named as=20 >=20 > [PATCH for 2.5 v6 0/10] dataplane snapshot fixes=20 >=20 > Den=20 anyway, even if above will not help, can you collect gdb=20 traces from all threads in QEMU process. May be I'll be=20 able to give a hit.=20 Den=20