(原)KVM环境下虚拟机迁移失败问题解决

使用开源云工具OpenNebula3.8.1在KVM环境下虚拟机迁移失败问题解决。
1、虚拟机迁移失败1日志：

Fri Mar 8 17:57:18 2013 [LCM][I]: New VM state is SAVE_MIGRATE

Fri Mar 8 17:57:30 2013 [VMM][I]: ExitCode: 0

Fri Mar 8 17:57:30 2013 [VMM][I]: Successfully execute virtualization driver operation: save.

Fri Mar 8 17:57:30 2013 [VMM][I]: ExitCode: 0

Fri Mar 8 17:57:30 2013 [VMM][I]: Successfully execute network driver operation: clean.

Fri Mar 8 17:58:14 2013 [LCM][I]: New VM state is PROLOG_MIGRATE

Fri Mar 8 17:58:14 2013 [TM][I]: mv: -------------------------/one_images_3.8.1/0/42/disk.0

Fri Mar 8 17:58:14 2013 [TM][I]: ExitCode: 0

Fri Mar 8 18:02:28 2013 [TM][I]: mv: Moving bcec162:/one_images_3.8.1/0/42 to node153:/one_images_3.8.1/0/42

Fri Mar 8 18:02:28 2013 [TM][I]: ExitCode: 0

Fri Mar 8 18:02:29 2013 [LCM][I]: New VM state is BOOT

Fri Mar 8 18:02:30 2013 [VMM][I]: ExitCode: 0

Fri Mar 8 18:02:30 2013 [VMM][I]: Successfully execute network driver operation: pre.

Fri Mar 8 18:02:33 2013 [VMM][I]: Command execution fail: /var/tmp/one/vmm/kvm/restore /one_images_3.8.1/0/42/checkpoint node153 42 node153

Fri Mar 8 18:02:33 2013 [VMM][E]: restore: Command "virsh --connect qemu:///system restore /one_images_3.8.1/0/42/checkpoint" failed: error: Failed to restore domain from /one_images_3.8.1/0/42/checkpoint

Fri Mar 8 18:02:33 2013 [VMM][I]: error: unable to set user and group to '0:0' on '/one_images_3.8.1/0/42/disk.1': No such file or directory

Fri Mar 8 18:02:33 2013 [VMM][E]: Could not restore from /one_images_3.8.1/0/42/checkpoint

Fri Mar 8 18:02:33 2013 [VMM][I]: ExitCode: 1

Fri Mar 8 18:02:33 2013 [VMM][I]: Failed to execute virtualization driver operation: restore.

Fri Mar 8 18:02:33 2013 [VMM][E]: Error restoring VM: Could not restore from /one_images_3.8.1/0/42/checkpoint

Fri Mar 8 18:02:34 2013 [DiM][I]: New VM state is FAILED

Sat Mar 9 09:23:46 2013 [DiM][I]: New VM state is DONE.

Sat Mar 9 09:23:46 2013 [TM][W]: Ignored: LOG I 42 ExitCode: 0

Sat Mar 9 09:23:47 2013 [TM][W]: Ignored: LOG I 42 delete: Deleting /one_images_3.8.1/0/42

Sat Mar 9 09:23:47 2013 [TM][W]: Ignored: LOG I 42 ExitCode: 0

Sat Mar 9 09:23:47 2013 [TM][W]: Ignored: TRANSFER SUCCESS 42 -

解决方法：

在mv脚本中TAR拷贝命令前面增加$SUDO命令.

$ONE_LOCATION/var/remotes/tm/ssh/mv

#!/bin/bash

# -------------------------------------------------------------------------- #

# #

# Licensed under the Apache License, Version 2.0 (the "License"); you may #

# not use this file except in compliance with the License. You may obtain #

# a copy of the License at #

# #

# http://www.apache.org/licenses/LICENSE-2.0 #

# #

# Unless required by applicable law or agreed to in writing, software #

# distributed under the License is distributed on an "AS IS" BASIS, #

# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. #

# See the License for the specific language governing permissions and #

# limitations under the License. #

#--------------------------------------------------------------------------- #

# MV <hostA:system_ds/disk.i|hostB:system_ds/disk.i> vmid dsid

# <hostA:system_ds/|hostB:system_ds/>

# - hostX is the target host to deploy the VM

# - system_ds is the path for the system datastore in the host

# - vmid is the id of the VM

# - dsid is the target datastore (0 is the system datastore)

SRC=$1

DST=$2

VMID=$3

DSID=$4

if [ -z "${ONE_LOCATION}" ]; then

TMCOMMON=/var/lib/one/remotes/tm/tm_common.sh

else

TMCOMMON=$ONE_LOCATION/var/remotes/tm/tm_common.sh

. $TMCOMMON

#-------------------------------------------------------------------------------

# Return if moving a disk, we will move them when moving the whole system_ds

# directory for the VM

#-------------------------------------------------------------------------------

SRC=`fix_dir_slashes $SRC`

DST=`fix_dir_slashes $DST`

SRC_PATH=`arg_path $SRC`

DST_PATH=`arg_path $DST`

SRC_HOST=`arg_host $SRC`

DST_HOST=`arg_host $DST`

DST_DIR=`dirname $DST_PATH`

SRC_DS_DIR=`dirname $SRC_PATH`

SRC_VM_DIR=`basename $SRC_PATH`

if [ `is_disk $DST_PATH` -eq 1 ]; then

log "-------------------------$DST_PATH"

exit 0

if [ "$SRC" == "$DST" ]; then

exit 0

ssh_make_path "$DST_HOST" "$DST_DIR"

log "Moving $SRC to $DST"

ssh_exec_and_log "$DST_HOST" "rm -rf '$DST_PATH'" \

"Error removing target path to prevent overwrite errors"

TAR_COPY="$SSH $SRC_HOST '$SUDO $TAR -C $SRC_DS_DIR -cf - $SRC_VM_DIR'"

TAR_COPY="$TAR_COPY | $SSH $DST_HOST '$TAR -C $DST_DIR -xf -'"

exec_and_log "eval $TAR_COPY" "Error copying disk directory to target host"

exec_and_log "$SSH $SRC_HOST rm -rf $SRC_PATH"

exit 0

-------------------------------------------------------------------------------------------

2、虚拟机迁移失败2日志：

Sat Mar 9 09:34:12 2013 [LCM][I]: New VM state is SAVE_MIGRATE

Sat Mar 9 09:34:24 2013 [VMM][I]: ExitCode: 0

Sat Mar 9 09:34:24 2013 [VMM][I]: Successfully execute virtualization driver operation: save.

Sat Mar 9 09:34:24 2013 [VMM][I]: ExitCode: 0

Sat Mar 9 09:34:24 2013 [VMM][I]: Successfully execute network driver operation: clean.

Sat Mar 9 09:34:25 2013 [LCM][I]: New VM state is PROLOG_MIGRATE

Sat Mar 9 09:34:25 2013 [TM][I]: mv: -------------------------/one_images_3.8.1/0/43/disk.0

Sat Mar 9 09:34:25 2013 [TM][I]: ExitCode: 0

Sat Mar 9 09:36:38 2013 [TM][I]: mv: Moving node153:/one_images_3.8.1/0/43 to bcec162:/one_images_3.8.1/0/43

Sat Mar 9 09:36:38 2013 [TM][I]: mv: -------------------target copyy

Sat Mar 9 09:36:38 2013 [TM][I]: mv: ++++++++++++++++++++++end copy

Sat Mar 9 09:36:38 2013 [TM][I]: ExitCode: 0

Sat Mar 9 09:36:38 2013 [LCM][I]: New VM state is BOOT

Sat Mar 9 09:36:38 2013 [VMM][I]: ExitCode: 0

Sat Mar 9 09:36:38 2013 [VMM][I]: Successfully execute network driver operation: pre.

Sat Mar 9 09:36:42 2013 [VMM][I]: Command execution fail: /var/tmp/one/vmm/kvm/restore /one_images_3.8.1/0/43/checkpoint bcec162 43 bcec162

Sat Mar 9 09:36:42 2013 [VMM][E]: restore: Command "virsh --connect qemu:///system restore /one_images_3.8.1/0/43/checkpoint" failed: error: Failed to restore domain from /one_images_3.8.1/0/43/checkpoint

Sat Mar 9 09:36:42 2013 [VMM][I]: error: internal error process exited while connecting to monitor: Supported machines are:

Sat Mar 9 09:36:42 2013 [VMM][I]: pc RHEL 6.0.0 PC (alias of rhel6.0.0)

Sat Mar 9 09:36:42 2013 [VMM][I]: rhel6.0.0 RHEL 6.0.0 PC (default)

Sat Mar 9 09:36:42 2013 [VMM][I]: rhel5.5.0 RHEL 5.5.0 PC

Sat Mar 9 09:36:42 2013 [VMM][I]: rhel5.4.4 RHEL 5.4.4 PC

Sat Mar 9 09:36:42 2013 [VMM][I]: rhel5.4.0 RHEL 5.4.0 PC

Sat Mar 9 09:36:42 2013 [VMM][E]: Could not restore from /one_images_3.8.1/0/43/checkpoint

Sat Mar 9 09:36:42 2013 [VMM][I]: ExitCode: 1

Sat Mar 9 09:36:42 2013 [VMM][I]: Failed to execute virtualization driver operation: restore.

Sat Mar 9 09:36:42 2013 [VMM][E]: Error restoring VM: Could not restore from /one_images_3.8.1/0/43/checkpoint

Sat Mar 9 09:36:42 2013 [DiM][I]: New VM state is FAILED

登陆到节点：

[root@bcec162 43]# virsh restore checkpoint

错误：从 checkpoint 恢复域失败

错误：internal error process exited while connecting to monitor: Supported machines are:

pc RHEL 6.0.0 PC (alias of rhel6.0.0)

rhel6.0.0 RHEL 6.0.0 PC (default)

rhel5.5.0 RHEL 5.5.0 PC

rhel5.4.4 RHEL 5.4.4 PC

rhel5.4.0 RHEL 5.4.0 PC

修改了bcec162节点的/etc/libvirt/qemu.conf文件:

# The user ID for QEMU processes run by the system instance

user = "root"

# The group ID for QEMU processes run by the system instance

group = "root"

# Whether libvirt should dynamically change file ownership

# to match the configured user/group above. Defaults to 1.

# Set to 0 to disable file ownership changes.

#dynamic_ownership = 0

bcec162节点迁移到node153节点成功。

[root@node153 43]# ll

total 5075464

-rw-r--r-- 1 root root 287215779 Mar 8 11:11 checkpoint

-rw-r--r-- 1 oneadmin kvm 283538737 Mar 9 09:34 checkpoint.1362712278

-rw-r--r-- 1 oneadmin kvm 920 Mar 9 09:26 deployment.0

-rw-r--r-- 1 root root 4621008896 Mar 9 10:14 disk.0

-rw-r----- 1 root root 401408 Mar 9 09:26 disk.1

lrwxrwxrwx 1 oneadmin kvm 29 Mar 9 10:09 disk.1.iso -> /one_images_3.8.1/0/43/disk.1

--------------------------------------------------------------------------------------------------------

3、仅修改node152节点的/etc/libvirt/qemu.conf文件：

# The user ID for QEMU processes run by the system instance

#user = "root"

# The group ID for QEMU processes run by the system instance

#group = "root"

# Whether libvirt should dynamically change file ownership

# to match the configured user/group above. Defaults to 1.

# Set to 0 to disable file ownership changes.

dynamic_ownership = 0

从bcec162节点迁移到node152不成功，日志如下：

Sat Mar 9 10:31:47 2013 [LCM][I]: New VM state is SAVE_MIGRATE

Sat Mar 9 10:31:54 2013 [VMM][I]: save: Moving old checkpoint file /one_images_3.8.1/0/43/checkpoint

Sat Mar 9 10:31:54 2013 [VMM][I]: ExitCode: 0

Sat Mar 9 10:31:54 2013 [VMM][I]: Successfully execute virtualization driver operation: save.

Sat Mar 9 10:31:54 2013 [VMM][I]: ExitCode: 0

Sat Mar 9 10:31:54 2013 [VMM][I]: Successfully execute network driver operation: clean.

Sat Mar 9 10:31:55 2013 [LCM][I]: New VM state is PROLOG_MIGRATE

Sat Mar 9 10:31:55 2013 [TM][I]: mv: -------------------------/one_images_3.8.1/0/43/disk.0

Sat Mar 9 10:31:55 2013 [TM][I]: ExitCode: 0

Sat Mar 9 10:35:02 2013 [TM][I]: mv: Moving bcec162:/one_images_3.8.1/0/43 to node152:/one_images_3.8.1/0/43

Sat Mar 9 10:35:02 2013 [TM][I]: mv: -------------------target copyy

Sat Mar 9 10:35:02 2013 [TM][I]: mv: ++++++++++++++++++++++end copy

Sat Mar 9 10:35:02 2013 [TM][I]: ExitCode: 0

Sat Mar 9 10:35:02 2013 [LCM][I]: New VM state is BOOT

Sat Mar 9 10:35:03 2013 [VMM][I]: ExitCode: 0

Sat Mar 9 10:35:03 2013 [VMM][I]: Successfully execute network driver operation: pre.

Sat Mar 9 10:35:07 2013 [VMM][I]: Command execution fail: /var/tmp/one/vmm/kvm/restore /one_images_3.8.1/0/43/checkpoint node152 43 node152

Sat Mar 9 10:35:07 2013 [VMM][E]: restore: Command "virsh --connect qemu:///system restore /one_images_3.8.1/0/43/checkpoint" failed: error: Failed to restore domain from /one_images_3.8.1/0/43/checkpoint

Sat Mar 9 10:35:07 2013 [VMM][I]: error: operation failed: failed to retrieve chardev info in qemu with 'info chardev'

Sat Mar 9 10:35:07 2013 [VMM][E]: Could not restore from /one_images_3.8.1/0/43/checkpoint

Sat Mar 9 10:35:07 2013 [VMM][I]: ExitCode: 1

Sat Mar 9 10:35:07 2013 [VMM][I]: Failed to execute virtualization driver operation: restore.

Sat Mar 9 10:35:07 2013 [VMM][E]: Error restoring VM: Could not restore from /one_images_3.8.1/0/43/checkpoint

Sat Mar 9 10:35:07 2013 [DiM][I]: New VM state is FAILED

登陆到node152节点执行restore命令：

[root@node152 43]# virsh restore checkpoint

error: Failed to restore domain from checkpoint

error: internal error process exited while connecting to monitor: qemu: could not open disk image /one_images_3.8.1/0/43/disk.0: Permission denied

将/etc/libvirt/qemu.conf文件中注释掉dynamic_ownership=0，开启user=root和group=root.

如果开启dynamic_ownership则恢复虚拟机也会报出上面的错误信息。

在node152节点上恢复虚拟机：

[root@node152 43]# virsh restore checkpoint

Domain restored from checkpoint

[root@node152 43]# virsh list

Id Name State

----------------------------------

117 one-43 running

参考文章：

https://wiki.archlinux.org/index.php/QEMU_(%E7%AE%80%E4%BD%93%E4%B8%AD%E6%96%87)

http://hi.baidu.com/juacm/item/f1fc3f98d8428ad07a7f01e2

转载请保持原链接：http://www.blogjava.net/ldwblog/archive/2013/03/08/396187.html

(原)KVM环境下虚拟机迁移失败问题解决

你可能感兴趣的:((原)KVM环境下虚拟机迁移失败问题解决)