[Users] (potential bug) After I create a virtual disk I am unable to start a VM - Permissions issue ? The node seems to have no write access to the directory on the NFS server...

Wed Mar 28 13:51:41 UTC 2012

On 03/28/2012 09:30 AM, Morgan Cox wrote:
> Hi.
>
> My setup is 3 servers
>
> 1. Frontend (engine).
> 2 Ovirt node
> 3. Nfs server (debian)
>
>  From the Frontend (engine) :-
>
> -bash-4.2$ ls -la
> /tmp/test/2f7ee7bc-09b3-42ba-af91-40d79293e360/images/d1fcc1ae-dcf5-426e-a99f-3a48e84c5ae3/
> total 2097168
> drwxr-xr-x. 2 vdsm kvm       4096 Mar 28  2012 .
> drwxr-xr-x. 3 vdsm kvm       4096 Mar 28  2012 ..
> -rw-rw----. 1 vdsm kvm 2147483648 Mar 28  2012
> 9fe193c9-7139-4a6c-933a-c6f31d5e96bd
> -rw-r--r--. 1 vdsm kvm        317 Mar 28  2012
> 9fe193c9-7139-4a6c-933a-c6f31d5e96bd.meta
>
>
> I have (now) write access with vdsm user from here
>
>  From the node (look at ownership....) :-
>
>
> -bash-4.2$ ls -la
> /rhev/data-center/b2b2e054-66b2-11e1-bda3-1728f784de9e/2f7ee7bc-09b3-42ba-af91-40d79293e360/images/d1fcc1ae-dcf5-426e-a99f-3a48e84c5ae3/
> total 2097168
> drwxr-xr-x. 2 4294967294 4294967294       4096 Mar 28 13:21 .
> drwxr-xr-x. 3 4294967294 4294967294       4096 Mar 28 13:05 ..
> -rw-rw----. 1 4294967294 4294967294 2147483648 Mar 28 13:05
> 9fe193c9-7139-4a6c-933a-c6f31d5e96bd
> -rw-r--r--. 1 4294967294 4294967294        317 Mar 28 13:05
> 9fe193c9-7139-4a6c-933a-c6f31d5e96bd.meta
>
Morgan,
Look at the UID/GID here.  They are definitely not 36:36; hence, the 
user VDSM *might* have an issue with R/W privileged from the node.

Can the user VDSM R/W anything in the mounted directory from the node?  
Also, if you don't want to create a UID/GID combination on the NFS 
server of 36:36 try pinning it...
$ cat /etc/exports
/virt/iso                      
192.168.122.11(rw,sync,all_squash,anonuid=107,anongid=107) <-- I'm 
pinning to 107 because I don't want 36:36 on my NFS server.

> Also
>
>  From the engine - mount command:-
>
> 10.0.0.190:/storage1/ on /tmp/test type nfs
> (rw,relatime,vers=3,rsize=1048576,wsize=1048576,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=sys,mountaddr=10.0.0.190,mountvers=3,mountport=35065,mountproto=udp,local_lock=none,addr=10.0.0.190)
>
> - note that the NFS share wasn't mounted until I mounted it...
>
>  From the node - mount command
>
> 10.0.0.190:/storage1/ on /rhev/data-center/mnt/10.0.0.190:_storage1
> type nfs4 (rw,relatime,vers=4,rsize=1048576,wsize=1048576,namlen=255,soft,nosharecache,proto=tcp,port=0,timeo=600,retrans=6,sec=sys,clientaddr=10.0.0.101,minorversion=0,local_lock=none,addr=10.0.0.190)
>
> - looks like nfs4 on the node........
>
> Any ideas anyone ?
>
> Cheers
>
>
>
>
>
>
> On 28 March 2012 06:45, Deepak C Shetty<deepakcs at linux.vnet.ibm.com>  wrote:
>> On 03/27/2012 04:08 PM, Morgan Cox wrote:
>>
>> Hi
>>
>> http://www.ovirt.org/wiki/Troubleshooting_NFS_Storage_Issues
>>
>> - This seems to refer to setup on a redhat based system
>>
>> The fies it refers to do not exist on Debian - i.e
>>
>> /etc/sysconfig/nfs and/etc/nfsmount.conf - they do not exist...
>>
>> Do you know the equivalent on Debian ?
>>
>>
>> Sorry, i use fedora, and I dont know the equivalent of Debain.
>> Maybe asking the big brain(google) might help. There should be an
>> equivalent way of doing the same on debian.
>>
>> Also in my case v4 was not a issue, as much as the perms and uid:gid
>> were, so look further to see perms and selinux booleans are
>> set correctly in your setup.
>>
>>
>> Regards
>>
>>
>> On 27 March 2012 07:07, Deepak C Shetty<deepakcs at linux.vnet.ibm.com>  wrote:
>>> On 03/26/2012 05:28 PM, Morgan Cox wrote:
>>>
>>> Hi.
>>>
>>> Still not actually managed to test a vm... I can start a VM (and use
>>> spice) without a virtual disk.
>>>
>>> However as soon as I add a virtual disk the VM no longer starts
>>>
>>>
>>> Go thru this and see if it helps...
>>> http://www.ovirt.org/wiki/Troubleshooting_NFS_Storage_Issues
>>>
>>>
>>> example error message
>>>
>>>
>>> -------------------------------------------------------------------------------------------------------------------
>>> VM ddf is down. Exit message internal error process exited while
>>> connecting to monitor: qemu-kvm: -drive
>>> file=/rhev/data-center/b2b2e054-66b2-11e1-bda3-1728f784de9e/2f7ee7bc-09b3-42ba-af91-40d79293e360/images/c507a2bc-38c5-498e-88c0-7cce8169cf67/6c36c8e4-6618-42f3-9cc4-06d2bccdc9cf,if=none,id=drive-virtio-disk0,format=raw,serial=c507a2bc-38c5-498e-88c0-7cce8169cf67,cache=none,werror=stop,rerror=stop,aio=threads:
>>> could not open disk image
>>> /rhev/data-center/b2b2e054-66b2-11e1-bda3-1728f784de9e/2f7ee7bc-09b3-42ba-af91-40d79293e360/images/c507a2bc-38c5-498e-88c0-7cce8169cf67/6c36c8e4-6618-42f3-9cc4-06d2bccdc9cf:
>>> Permission denied .
>>>
>>> -------------------------------------------------------------------------------------------------------------------
>>>
>>> I have noticed that the
>>> directory /rhev/data-center/b2b2e054-66b2-11e1-bda3-1728f784de9e/ is not
>>> writable via the ovirt node server.
>>>
>>> I am using a separate NFS server for storage.
>>>
>>> Is this a bug ?
>>>
>>> Does anyone know how to fix this ?
>>>
>>> shall I report it ?
>>>
>>> Many regards !
>>>
>>>
>>> _______________________________________________
>>> Users mailing list
>>> Users at ovirt.org
>>> http://lists.ovirt.org/mailman/listinfo/users
>>>
>>>
>>
> _______________________________________________
> Users mailing list
> Users at ovirt.org
> http://lists.ovirt.org/mailman/listinfo/users