[one-users] can't create VM with shared vm dir

[Shantanu Pavgi]

Hi,

I have configured OpenNebula to use shared NFS for image repo and VM_DIR, e.g. /groups/one-vm-dir and /groups/one-images-dir are exported from a separate NFS server, both head-node and cluster-nodes have mounted these shares. Following is OpenNebula config:

{{{
DB=BACKEND=sqlite
DEBUG_LEVEL=3
DEFAULT_DEVICE_PREFIX=hd
DEFAULT_IMAGE_TYPE=OS
HM_MAD=EXECUTABLE=one_hm
HOST_MONITORING_INTERVAL=600
IMAGE_REPOSITORY_PATH=/groups/one-images-dir
IM_MAD=ARGUMENTS=kvm,EXECUTABLE=one_im_ssh,NAME=im_kvm
MAC_PREFIX=02:00
MANAGER_TIMER=15
NETWORK_SIZE=254
PORT=2633
SCRIPTS_REMOTE_DIR=/var/tmp/one
TM_MAD=ARGUMENTS=tm_nfs/tm_nfs.conf,EXECUTABLE=one_tm,NAME=tm_nfs
VM_DIR=/groups/one-vm-dir
VM_HOOK=ARGUMENTS=$VMID,COMMAND=image.rb,NAME=image,ON=DONE
VM_MAD=ARGUMENTS=kvm,DEFAULT=vmm_ssh/vmm_ssh_kvm.conf,EXECUTABLE=one_vmm_ssh,NAME=vmm_kvm,TYPE=kvm
VM_POLLING_INTERVAL=600
VNC_BASE_PORT=5900
}}}


I am getting following error after creating a VM. The TM is creating disk.0, deployment.0 and log files in /srv/cloud/one/var ($ONE_LOCATION/var). The VM dir (/groups/one-vm-dir) contains only deployment.0 file and disk.0 is not created here. The VMM is failing to create a VM as there is no disk image in the VM_DIR. 

{{{
Fri Mar  4 12:50:23 2011 [DiM][I]: New VM state is ACTIVE.
Fri Mar  4 12:50:24 2011 [LCM][I]: New VM state is PROLOG.
Fri Mar  4 12:50:24 2011 [VM][I]: Virtual Machine has no context
Fri Mar  4 12:50:32 2011 [TM][I]: tm_clone.sh: opennebula.exp.uab.edu:/groups/one-images-dir/ttylinux/ttylinux.img kvm.exp.uab.edu:/groups/one-vm-dir/16/images/disk.0
Fri Mar  4 12:50:32 2011 [TM][I]: tm_clone.sh: DST: /srv/cloud/one/var/16/images/disk.0
Fri Mar  4 12:50:32 2011 [TM][I]: tm_clone.sh: Creating directory /srv/cloud/one/var/16/images
Fri Mar  4 12:50:32 2011 [TM][I]: tm_clone.sh: Executed "mkdir -p /srv/cloud/one/var/16/images".
Fri Mar  4 12:50:32 2011 [TM][I]: tm_clone.sh: Executed "chmod a+w /srv/cloud/one/var/16/images".
Fri Mar  4 12:50:32 2011 [TM][I]: tm_clone.sh: Cloning /groups/one-images-dir/ttylinux/ttylinux.img
Fri Mar  4 12:50:32 2011 [TM][I]: tm_clone.sh: Executed "cp -r /groups/one-images-dir/ttylinux/ttylinux.img /srv/cloud/one/var/16/images/disk.0".
Fri Mar  4 12:50:32 2011 [TM][I]: tm_clone.sh: Executed "chmod a+rw /srv/cloud/one/var/16/images/disk.0".
Fri Mar  4 12:50:33 2011 [LCM][I]: New VM state is BOOT
Fri Mar  4 12:50:33 2011 [VMM][I]: Generating deployment file: /srv/cloud/one/var/16/deployment.0
Fri Mar  4 12:51:07 2011 [VMM][I]: Command execution fail: 'if [ -x "/var/tmp/one/vmm/kvm/deploy" ]; then /var/tmp/one/vmm/kvm/deploy /groups/one-vm-dir/16/images/deployment.0; else                              exit 42; fi'
Fri Mar  4 12:51:07 2011 [VMM][I]: STDERR follows.
Fri Mar  4 12:51:07 2011 [VMM][I]: error: Failed to create domain from /groups/one-vm-dir/16/images/deployment.0
Fri Mar  4 12:51:07 2011 [VMM][I]: error: internal error unable to start guest: qemu: could not open disk image /groups/one-vm-dir/16/images/disk.0
Fri Mar  4 12:51:07 2011 [VMM][I]: 
Fri Mar  4 12:51:07 2011 [VMM][I]: ExitCode: 255
Fri Mar  4 12:51:07 2011 [VMM][E]: Error deploying virtual machine: error: Failed to create domain from /groups/one-vm-dir/16/images/deployment.0
Fri Mar  4 12:51:08 2011 [DiM][I]: New VM state is FAILED
Fri Mar  4 12:51:08 2011 [TM][W]: Ignored: LOG - 16 tm_delete.sh: Deleting /srv/cloud/one/var/16/images
Fri Mar  4 12:51:09 2011 [TM][W]: Ignored: LOG - 16 tm_delete.sh: Executed "rm -rf /srv/cloud/one/var/16/images".
}}}


Am I missing something in the configuration here? Any pointers?


Thanks,
Shantanu.