VMware Cloud Community
ckoeber
Contributor
Contributor

NFS Configuration Problem, Unable to connect from VMWare Server

Hello,

I am trying to add NFS storage to an ESX 3.5 server. The folloing is what I have done so far.

NFS Server:

  • On a class C network: 10.10.10.X

    • IP Address of NFS FileServer: 10.10.10.105

    • Subnet mask: 255.255.255.0

    • Gateway: 10.10.10.1

  • /etc/exports file looks like this:

    • /wtsdata 10.10.1.1/24(rw,insecure,sync,no_root_squash) 10.10.10.1/24(rw,insecure,sync,no_root_squash)

  • Firewall has been disabled

  • Using Red Hat Enterprise Linux 5 (64-bit version

  • I can mount the system on another unrelated linux installation using:

    • mount -t nfs 10.10.10.105:/wtsdata /mountpoint

    • The system used to test is on the same subnet

VMWare ESX 3,5 Server:

  • On a class C network: 10.10.1.X

    • IP Address of ESX 3.5 Server (VM Kernel Port): 10.10.1.40

    • Subnet Mask: 255.255.255.0

    • Gateway: 10.10.1.1

    • Vmotion Enabled

  • Firewall Configuration edited for NFS Client

I can ping the ESX 3.5 server from the NFS file server. However, whenever I try to add storage to the ESX 3.5 server and select the Network File System option (filling out all of the fields, of course) the console client that I use returns the following error message:

_ "Error during the configuration of the host: NFS Error: Unable to Mount filesystem: Unable to connect to NFS server"_

So, my question is what am I doing wrong? I presume this is a simple setup but, of course, I may be naive.

Thank you for your time.

Regards,

Chris K.

Reply
0 Kudos
10 Replies
fejf
Expert
Expert

I can mount the system on another unrelated linux installation using:

Was this Linux box in the 10.10.1.x subnet? Or in other words: did you check if the router from the 10.10.1.x/24 to the 10.10.10.x/24 network allows nfs to pass through? Have you check if your Linux Client uses NFSv3 with TCP (not NFSv4 and not UDP) ?

--

There are 10 types of people. Those who understand binary and the rest. And those who understand gray-code.

-- There are 10 types of people. Those who understand binary and the rest. And those who understand gray-code.
ckoeber
Contributor
Contributor

Was this Linux box in the 10.10.1.x subnet? Or in other words: did you check if the router from the 10.10.1.x/24 to the 10.10.10.x/24 network allows nfs to pass through? Have you check if your Linux Client uses NFSv3 with TCP (not NFSv4 and not UDP) ?

Both the unrelated linux box and the NFS server were on the same subnet at the time BUT no traffic is blocked from the two subnets. (We have a Windows Domain with nodes in various subnets.)

Thanks for your time.

Reply
0 Kudos
weinstein5
Immortal
Immortal

Welcome to the forums - From the service console can you ping form the vmkernel port using the vmkping command - the syntax vmkping 10.10.10.105? Can you ping the gateway address of 10.10.1.1 using the vmkping command? Can you ping the vmkernel port form your NFS server? - I am thinking it is an issue of communication between you vmkernel port and your NFS server -

Also you do not need to open the NFS port on the firewall to connect to an NFS datastore -

If you find this or any other answer useful please consider awarding points by marking the answer correct or useful

If you find this or any other answer useful please consider awarding points by marking the answer correct or helpful
ckoeber
Contributor
Contributor

Welcome to the forums - From the service console can you ping form the vmkernel port using the vmkping command - the syntax vmkping 10.10.10.105? Can you ping the gateway address of 10.10.1.1 using the vmkping command? Can you ping the vmkernel port form your NFS server? - I am thinking it is an issue of communication between you vmkernel port and your NFS server -

Also you do not need to open the NFS port on the firewall to connect to an NFS datastore -

If you find this or any other answer useful please consider awarding points by marking the answer correct or useful

I have the answers to your questions!

  • I cannot ping ANY host from the vmkernel port (10.10.1.40) to any server - including the gateway - EXCEPT another vmware server (weird) ...

  • Other systems CAN ping the VMKernel port. (Including systems within the 10.10.1.X and 10.10.10.X subnets)

No traffice is blocked on any of the non-vmware systems (all firewalls on other systems are disabled as everything is on a local lan and protected by a hardware firewall that segregates the LAN from the WAN.

Any ideas?

Thanks for the time.

Reply
0 Kudos
ckoeber
Contributor
Contributor

Also, I should have version 4 support on the server. This is the output of the command I used to check:

rpcinfo -p localhost | grep -i nfs
Output:
100003 2 udp 2049 nfs
100003 3 udp 2049 nfs
100003 4 udp 2049 nfs
100003 2 tcp 2049 nfs
100003 3 tcp 2049 nfs
100003 4 tcp 2049 nfs

I can mount the system on another unrelated linux installation using:

Was this Linux box in the 10.10.1.x subnet? Or in other words: did you check if the router from the 10.10.1.x/24 to the 10.10.10.x/24 network allows nfs to pass through? Have you check if your Linux Client uses NFSv3 with TCP (not NFSv4 and not UDP) ?

--

There are 10 types of people. Those who understand binary and the rest. And those who understand gray-code.

Reply
0 Kudos
cowlesav
Contributor
Contributor

I am also experiencing a similar error. My situation differs a little bit as my NFS export is setup via a Netapp NAS. I have 3 ESX servers running 3.5, and my problem seems to be instability with the virtuals when the networking into the filer is full. I have added a 2nd link to the filer in hopes of isolating the vmware mounts from the rest of the network traffic but find that I can't mount nfs via this link. As with the solution above, I find that I cannot vmkping the new interface. I am trying to figure out how to open this. If anyone has any helpful suggestions I will be most appreciative.

Edited note: It's funny, I can mount off the other adapter via hostname on all these boxes, just not off the new one, both the old and new are on the same subnet, also I have noticed that even though I can mount the storage on the old link, I still cannot vmkping it. Is this normal?

AC

Reply
0 Kudos
Texiwill
Leadership
Leadership

Hello,

If you can not vmkping the NFS server then ESX can not reach the NFS Server. Remember, to use NFS you need a vmkernel port. You can not use the Service Console for this. But another vmkernel port just for NFS on a different subnet than SC, vMotion, etc.


Best regards,

Edward L. Haletky

VMware Communities User Moderator

====

Author of the book 'VMWare ESX Server in the Enterprise: Planning and Securing Virtualization Servers', Copyright 2008 Pearson Education.

CIO Virtualization Blog: http://www.cio.com/blog/index/topic/168354

As well as the Virtualization Wiki at http://www.astroarch.com/wiki/index.php/Virtualization

--
Edward L. Haletky
vExpert XIV: 2009-2023,
VMTN Community Moderator
vSphere Upgrade Saga: https://www.astroarch.com/blogs
GitHub Repo: https://github.com/Texiwill
Reply
0 Kudos
weinstein5
Immortal
Immortal

As Ed indicated if you can not ping via vmkping then the vmkernel port cannot communicate with your NFS server so it will not mount - at this point I would look at your physical network - making sure the gateway is functioning, any vlans are configured correctly, etc

If you find this or any other answer useful please consider awarding points by marking the answer correct or helpful
Reply
0 Kudos
AndrewSt
Enthusiast
Enthusiast

The ESX server - what is the vmkernel ip?? It looks like the problem is your exports file.

/etc/exports file looks like this:

  • /wtsdata 10.10.1.1/24(rw,insecure,sync,no_root_squash) 10.10.10.1/24(rw,insecure,sync,no_root_squash)

If you want the entire class C to be able to mount, shouldn't that be 10.10.1.0/24??

-


-Andrew Stueve

-Points for good answers are always appreciated

----------------------- -Andrew Stueve -Remember, if you found this or any other answer useful, please consider the use of the Helpful or Correct buttons to award points
Reply
0 Kudos
cowlesav
Contributor
Contributor

After rerunning network cables and only plugging up the adapters I needed I found I can now use the vmkping and regular ping from the service console, seemed to be an issue of misconfigured network on the part of our cisco guys. Thanks for the suggestions. I am now mounting all of my storage via a 10G fibre link.

AC

Reply
0 Kudos