Hello, I am trying to connect a NFS server (a debian 4 etch) with a ESX 3.5 update 1 and I get the following error:
"Error during the configuration of the host: NFS Error: Unable to Mount filesystem : Unable to connect to NFS server"
ESX host has a 192.168.18.2 / 255.255.0.0 ip configuration and NFS debian server is 192.168.15.36/255.255.0.0,
I have a virtual switch with a VMKernel and Service Console ip address inside this range (I attach the image)
I can make a vmkping fron the ESX to the NFS server.
ESX firewall for NFS clients is open
Debain NFS server is working fine, beacuse I have mounted the volumen in another machine for to be sure that the problem is not the nfs server. (ESX ip address and directory are on the /etc/exports) as
/var/nfs 192.168.18.2(rw,sync)
This is the log of the vmkernel
Jun 11 13:39:31 virtual vmkernel: 2:00:38:22.670 cpu2:1036)NFS: 107: Command: (mount) Server: (192.168.15.36) IP: (192.168.15.36) Path: (/var/nfs) Label: (PRUEBAS) Options: (None)
Jun 11 13:40:01 virtual vmkernel: 2:00:38:53.608 cpu2:1036)WARNING: NFS: 931: MOUNT RPC failed with RPC status 13 (RPC was aborted due to timeout) trying to mount Server (192.168.15.36) Path (/var/nfs)
It is like it can connect with 192.168.15.36 but they are in the same subnet .... and I can make vmkping ..
Any idea?
Thanks a lot
Do you have a gateway defined for the vmkernel port? Just noticed that the NFS server is on a different subnet form the vmkernle port so without a gateway defined you will not be able to communicate to the NFS server - and only the vmkernel port needs to see the nas/nfs device -
Also do you have no_root_squash configured for the folder you are sharing -
Try adding the NFS server to the ESX Host's /etc/hosts file. I've had mixed results using the IP from ESX hosts. Also try reloading or restarting the NFS services on the NFS server.
No, it does not works, I put the host in the /etc/hosts and reload nfs services on the NFS server, also try to put host on the /etc/hosts.allow and is doest not work too.
Thanks a lot
You may want to try mounting the NFS server from the service console. Make sure you start the portmapper and nfs:
service portmap start
service nfs start
Then try to mount the NFS server and see what errors get produced. Also check /var/log/messages in the service console.
On the NFS server make sure you are not getting killed by SELINUX. /var/log/messages on the NFS server should help also.
Yes!!, now, I can mount the NFS server from the service consele using
mount 192.168.15.36:/var/nfs /mnt/cdrom
after start nfs and portmap
BUT !!!, I cant do it from the GUI ESX ... It show me the same error !!
I am not positive but I think the service console and the vmkernel port have to be on different subnets, but both need to be able to talk to the nfs server.
Do you have a gateway defined for the vmkernel port? Just noticed that the NFS server is on a different subnet form the vmkernle port so without a gateway defined you will not be able to communicate to the NFS server - and only the vmkernel port needs to see the nas/nfs device -
Also do you have no_root_squash configured for the folder you are sharing -
For the vmkernel to connect to a nas/nfs device you do not need to open the NFS port on the service console firewall -
if you are still having issues with the mount in the GUI, add the vmkernel IP address to your export on the NFS server. It is the vmkernel interface that mounts to the NFS share for a datastore mount.
Yes, this was the problem, I was trying with the ip address of the ESX server, not the vmkernel, on the debian NFS /etc/exports
Now,it works fine.
Thanks
Hi Roluje,
One more doube on this.
Why it is not resolving the name??
Regards,
Aravind