From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mx1.redhat.com (ext-mx02.extmail.prod.ext.phx2.redhat.com [10.5.110.26]) by int-mx14.intmail.prod.int.phx2.redhat.com (8.14.4/8.14.4) with ESMTP id t2LN8LQm007818 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-GCM-SHA384 bits=256 verify=NO) for ; Sat, 21 Mar 2015 19:08:21 -0400 Received: from esa2.ucsf.iphmx.com (unknown [68.232.143.34]) by mx1.redhat.com (Postfix) with ESMTPS id 56FAAB6A27 for ; Sat, 21 Mar 2015 23:08:14 +0000 (UTC) Received: from exht05.net.ucsf.edu (mx.ucsf.edu [64.54.247.193]) by bcuda1.ucsf.edu with ESMTP id qlbTSl3c8iw2s3mr (version=TLSv1 cipher=AES128-SHA bits=128 verify=NO) for ; Sat, 21 Mar 2015 16:08:03 -0700 (PDT) From: "Boylan, Ross" Date: Sat, 21 Mar 2015 23:08:03 +0000 Message-ID: Content-Language: en-US Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Subject: [linux-lvm] Questions about invalid snapshot Reply-To: LVM general discussion and development List-Id: LVM general discussion and development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , List-Id: To: "linux-lvm@redhat.com" The kernel logs show Mar 21 03:39:06 markov00 kernel: [809931.281437] device-mapper: snapshots: Invalidating snapshot: Unable to allocate exception. Mar 21 03:39:06 markov00 kernel: [809931.283835] Buffer I/O error on device dm-52, logical block 3882845 Mar 21 03:39:06 markov00 kernel: [809931.283838] lost page write due to I/O error on dm-52 Mar 21 03:39:06 markov00 kernel: [809931.283840] Buffer I/O error on device dm-52, logical block 3882811 Mar 21 03:39:06 markov00 kernel: [809931.283843] lost page write due to I/O error on dm-52 # and many more such messages Mar 21 03:39:33 markov00 kernel: [809957.940036] lost page write due to I/O error on dm-52 Mar 21 03:39:36 markov00 kernel: [809961.553242] sched: RT throttling activated Mar 21 03:39:37 markov00 kernel: [809962.062074] Buffer I/O error on device dm-52, logical block 256 Mar 21 03:39:37 markov00 kernel: [809962.062416] Buffer I/O error on device dm-52, logical block 256 Mar 21 03:39:37 markov00 kernel: [809962.063086] Buffer I/O error on device dm-52, logical block 256 Mar 21 03:39:37 markov00 kernel: [809962.063406] Buffer I/O error on device dm-52, logical block 256 Mar 21 03:39:37 markov00 kernel: [809962.063714] Buffer I/O error on device dm-52, logical block 256 and the snapshot volume is inactive. # lvchange -ay bigboy/capsdt02 File descriptor 3 (/dev/tty) leaked on lvchange invocation. Parent PID 6544: /bin/bash /dev/bigboy/capsdt02: read failed after 0 of 4096 at 182536044544: Input/output error /dev/bigboy/capsdt02: read failed after 0 of 4096 at 182536101888: Input/output error /dev/bigboy/capsdt02: read failed after 0 of 4096 at 0: Input/output error /dev/bigboy/capsdt02: read failed after 0 of 4096 at 4096: Input/output error Change of snapshot capsdt02 will also change its origin capsdt and 1 other snapshot(s). Proceed? [y/n]: n Logical volume capsdt02 not changed. 1) Is this from the snapshot filling, or something else going on? 2) Is it really true that changing the snapshot can change the origin and other snapshots? I thought that should be impossible. 3) Is there any way to recover the state of the snapshot at some point before it filled? The origin has been unchanged. 4) A recent thread on this lists suggests using thin provisioning for snapshots. man lvcreate seems to indicate that this is only possible if the origin is thin provisioned ("Thin snapshot is created when the origin is a thin volume and the size is not specified."). Is my understanding correct? In this case I have an LV capsdt that is a bit-image of the disk of a physical machine, and 2 snapshots, capsdt01 and capsdt02 based on it. I have run the snapshots as raw disks for 2 KVM virtual machines; the guest OS is Windows 7. Otherwise I have not touched capsdt. LVM 2.02.95 on Linux 3.2.0-4-amd64, Debian wheezy. Thanks. Ross Boylan