Intermittent "Error: (9, Bad File Descriptor)" Errors for all xm commands
Hi,
We are running several Oracle Enterprise Linux 32bit VMs on VM Server and now and again we will get into a state whereby any 'xm' command on the underlying physical server will start to give "Error: (9, Bad File Descriptor)" errors, even a simple 'xm list'.
I had come across this error before when issuing 'xm create' and accidentally putting the switches in the wrong order, but this time all I did was:
1. Power Off a running VM.
2. Created a template successfully from the VM using VM Manager.
3. Tried to restart the original VM in VM Manager and it said initialising for over 30 mins.
4. Checked the physical server and found that I couldn't use any xm commands due to the above File Descriptor error.
A typical vm.cfg file that we're using looks like this:
bootloader = '/usr/bin/pygrub'
disk = ['file:/OVS/running_pool/136_coesoa1/System.img,xvda,w',
',hdc:cdrom,r',
]
memory = '2048'
name = '136_coesoa1'
on_crash = 'restart'
on_reboot = 'restart'
vcpus = 2
vfb = ['type=vnc,vncunused=1']
vif = ['mac=00:16:3E:45:80:81, bridge=xenbr0']
Currently I am having to reboot the physical server to get xm to work again. Has anyone else come across this type of problem and/or is there any other way to reset xm so that it works again without having to reboot?
Many thanks.
Indy.