VMware 虚拟机无法从命令行恢复
我在 Ubuntu 10.10 上运行 VMWare Workstation。我每周都会克隆一系列虚拟机。
我编写了一个 bash 脚本,它循环遍历每个虚拟机,轻柔地挂起它,克隆它,然后尝试恢复它。
问题是,虚拟机没有恢复,所以我必须手动恢复它们。我正在尝试找出如何修改我编写的脚本以确保虚拟机恢复。我已经包含了脚本和生成的显示错误消息的日志文件的示例。
#!/bin/bash
##Author: William Cooper
##Date: 2/15/2011
## Purpose: Perform full backups of Virtual Machines
## Running under VMWare Workstation
## Script won't run if the /nas directory doesn't exist
## The /nas directory is a mounted Buffalo Terastation
## Check /etc/fstab if this mountpoint is broken
BACKUPDEST="/nas"
## No need to modify
HOST=`hostname`
DATEFILE=`/bin/date +%G%m%d`
## Run Command to find list of VM names:
## this will only be the name of the vm, no path and no .vmx
VMLIST=`vmrun list | grep '/' | cut -d'/' -f3 | cut -d'.' -f1`
## Initialize Counter variable used to cycle through VMs in VMARRAY variable
COUNTER=1
## Run Command to find list of VM Names and then store them in an Array:
## This will include full pathnames
VMARRAY=( `vmrun list | grep '/'` )
#################################################################
## Functions
##
## Notify the starting time of backup in the log
##
function startScript {
echo
echo "----------------------------------------------------"
echo "START - ${VM}"
echo "Host: ${HOST}"
echo "Date: `date`"
echo
}
##
## Suspend VM
##
function suspendVM {
## Suspend the VM
echo "Attempting to softly suspend "${VM}" . . ."
vmrun suspend ${VMARRAY[${COUNTER}]} soft
echo "${VM} Suspended on `date`"
echo
}
##
## Backup VM
##
function doBackup {
## Check to see if the correct backup directory exists.
if [ ! -d ${BACKUPDEST}/${VM} ]; then
echo "${BACKUPDEST}/${VM} does not exist, creating . . ."
mkdir "${BACKUPDEST}/${VM}"
echo
fi
## Backup VM (clone)
echo "Backup of "${VM}" began: `date`"
echo "Backup to ${BACKUPDEST}/${VM}/BACKUP/${DATEFILE} . . ."
mkdir ${BACKUPDEST}/${VM}/BACKUP/${DATEFILE}
vmrun clone ${VMARRAY[${COUNTER}]} ${BACKUPDEST}/${VM}/BACKUP/${DATEFILE}/${VM}.vmx full
echo "Backup of "${VM}" ended: `date`"
echo
}
##
## Start VM
##
function doStart {
## Resume VMs which were running [--type gui|sdl|vrdp|headless]
echo "Attempting to resume "${VM}" . . ."
vmrun start ${VMARRAY[${COUNTER}]}
echo "${VM} Resumed on `date`"
echo
}
##
## Notify the finishing time of backup in the log
##
function finishScript {
echo "FINISH - ${VM}"
echo "Host: ${HOST}"
echo "Date: `date`"
echo "----------------------------------------------------"
echo
}
#################################################################
## Script
for VM in ${VMLIST}
do
sleep 1
## StartScrip executes for logging
startScript
## Suspend VM performs soft suspend
suspendVM
sleep 3
## Do Backup performs the VM clone
doBackup
sleep 3
## Start VM performs start vm
doStart
sleep 3
## FinishScript executes for logging
finishScript
## Increment the Counter
COUNTER=${COUNTER}+1
done | tee "${BACKUPDEST}/logs/VMClone_${DATEFILE}.log"
## Script
#################################################################
exit
下面显示了为每个虚拟机生成的日志文件的示例。
----------------------------------------------------
START - vmname
Host: servername
Date: Sat Feb 19 12:30:29 CST 2011
Attempting to softly suspend vmname . . .
vmname Suspended on Sat Feb 19 12:30:44 CST 2011
Backup of vmname began: Sat Feb 19 12:30:47 CST 2011
Backup to /nas/vmname/BACKUP/20110219 . . .
Backup of vmname ended: Sat Feb 19 12:43:16 CST 2011
Attempting to resume vmname . . .
Error: Cannot launch the UI because no display server is present in the current environment
vmname Resumed on Sat Feb 19 12:43:20 CST 2011
FINISH - vmname
Host: servername
Date: Sat Feb 19 12:43:23 CST 2011
----------------------------------------------------
错误:无法启动 UI,因为当前环境中不存在显示服务器
I'm running VMWare Workstation on Ubuntu 10.10. I have a series of virtual machines that I clone every week.
I've written a bash script that cycles through each of the virtual machines, softly suspends it, clones it, and then attempts to resume it.
The problem is, the VM's do not resume, so I have to manually resume them. I'm trying to figure out how I can modify the script I've written to ensure the VMs resume. I've included the script and a sample of the generated logfile showing the error message.
#!/bin/bash
##Author: William Cooper
##Date: 2/15/2011
## Purpose: Perform full backups of Virtual Machines
## Running under VMWare Workstation
## Script won't run if the /nas directory doesn't exist
## The /nas directory is a mounted Buffalo Terastation
## Check /etc/fstab if this mountpoint is broken
BACKUPDEST="/nas"
## No need to modify
HOST=`hostname`
DATEFILE=`/bin/date +%G%m%d`
## Run Command to find list of VM names:
## this will only be the name of the vm, no path and no .vmx
VMLIST=`vmrun list | grep '/' | cut -d'/' -f3 | cut -d'.' -f1`
## Initialize Counter variable used to cycle through VMs in VMARRAY variable
COUNTER=1
## Run Command to find list of VM Names and then store them in an Array:
## This will include full pathnames
VMARRAY=( `vmrun list | grep '/'` )
#################################################################
## Functions
##
## Notify the starting time of backup in the log
##
function startScript {
echo
echo "----------------------------------------------------"
echo "START - ${VM}"
echo "Host: ${HOST}"
echo "Date: `date`"
echo
}
##
## Suspend VM
##
function suspendVM {
## Suspend the VM
echo "Attempting to softly suspend "${VM}" . . ."
vmrun suspend ${VMARRAY[${COUNTER}]} soft
echo "${VM} Suspended on `date`"
echo
}
##
## Backup VM
##
function doBackup {
## Check to see if the correct backup directory exists.
if [ ! -d ${BACKUPDEST}/${VM} ]; then
echo "${BACKUPDEST}/${VM} does not exist, creating . . ."
mkdir "${BACKUPDEST}/${VM}"
echo
fi
## Backup VM (clone)
echo "Backup of "${VM}" began: `date`"
echo "Backup to ${BACKUPDEST}/${VM}/BACKUP/${DATEFILE} . . ."
mkdir ${BACKUPDEST}/${VM}/BACKUP/${DATEFILE}
vmrun clone ${VMARRAY[${COUNTER}]} ${BACKUPDEST}/${VM}/BACKUP/${DATEFILE}/${VM}.vmx full
echo "Backup of "${VM}" ended: `date`"
echo
}
##
## Start VM
##
function doStart {
## Resume VMs which were running [--type gui|sdl|vrdp|headless]
echo "Attempting to resume "${VM}" . . ."
vmrun start ${VMARRAY[${COUNTER}]}
echo "${VM} Resumed on `date`"
echo
}
##
## Notify the finishing time of backup in the log
##
function finishScript {
echo "FINISH - ${VM}"
echo "Host: ${HOST}"
echo "Date: `date`"
echo "----------------------------------------------------"
echo
}
#################################################################
## Script
for VM in ${VMLIST}
do
sleep 1
## StartScrip executes for logging
startScript
## Suspend VM performs soft suspend
suspendVM
sleep 3
## Do Backup performs the VM clone
doBackup
sleep 3
## Start VM performs start vm
doStart
sleep 3
## FinishScript executes for logging
finishScript
## Increment the Counter
COUNTER=${COUNTER}+1
done | tee "${BACKUPDEST}/logs/VMClone_${DATEFILE}.log"
## Script
#################################################################
exit
An example of the generated logfile for each VM is shown below.
----------------------------------------------------
START - vmname
Host: servername
Date: Sat Feb 19 12:30:29 CST 2011
Attempting to softly suspend vmname . . .
vmname Suspended on Sat Feb 19 12:30:44 CST 2011
Backup of vmname began: Sat Feb 19 12:30:47 CST 2011
Backup to /nas/vmname/BACKUP/20110219 . . .
Backup of vmname ended: Sat Feb 19 12:43:16 CST 2011
Attempting to resume vmname . . .
Error: Cannot launch the UI because no display server is present in the current environment
vmname Resumed on Sat Feb 19 12:43:20 CST 2011
FINISH - vmname
Host: servername
Date: Sat Feb 19 12:43:23 CST 2011
----------------------------------------------------
Error: Cannot launch the UI because no display server is present in the current environment
如果你对这篇内容有疑问,欢迎到本站社区发帖提问 参与讨论,获取更多帮助,或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱,如果其他用户或者作者回复了您的评论,将不能在第一时间通知您!
发布评论
评论(4)
您可以在启动命令中使用 nogui 参数,例如:
请参阅:http://www.vmware.com/pdf/vix180_vmrun_command .pdf
You could use the nogui paramater to the start command, e.g:
See: http://www.vmware.com/pdf/vix180_vmrun_command.pdf
由于没有找到可靠的答案,我只能接受在这些特定情况下没有解决办法。
解决方案:关闭 VMware Workstation 并选择虚拟机“在后台运行”
之后工作完美无缺。
Since no solid answer has been found I'm just going to accept that under these specific circumstances there is no fix.
Solution: Close down VMware Workstation and choose for the VMs to "Run in the Background"
Works flawlessly after that.
尝试在运行 vmware 工作站的 TTY 终端上执行 echo $DISPLAY。
echo $DISPLAY 将返回:1000.0
你需要设置
在脚本中导出 DISPLAY=":1000.0" 。然后脚本将恢复虚拟机的..
当我使用槟榔备份来恢复我的虚拟机(使用在槟榔中运行的 postscript)时,我发现这有效。
try doing echo $DISPLAY on your terminal TTY where vmware workstation runs.
The echo $DISPLAY will return like :1000.0
you need to set
export DISPLAY=":1000.0" in your script. then the script will resume the VM's..
i found this works when i use areca backup to resume my VM using postscript run in areca.
如果问题是因为您从 ssh 会话运行 vmrun,则另一个解决方案可能是在服务器上运行 screen 会话。
在工作站正在运行的计算机上:
在客户端计算机上,
if the problem is because you are running vmrun from a ssh session, then another solution might be run a screen session on the server.
on the machine Workstation is running:
on the client machine,