VMware Solutions Discussions

question regarding VCS and NFS configuration

daniel_kaiser
3,149 Views

Hi All,

I have a vCenter server with VCS installed along with an ESX4 server configured to use NFS on a separate subnet.  The problem is when i try to provision storage using VCS it creates the NFS datastore but on the wrong subnet. I want it to create the datastore connecting to the separate NFS subnet but instead it's connecting to the production LAN subnet.  I assume this is happening because vCenter server itself knows only about the production subnet.  Any help would greatly be appreciated.

vCenter server is only connected to one subnet which is our production LAN.

ESX server itself is connected to multiple subnets including the dedicated NFS vlan.

Dan

1 ACCEPTED SOLUTION

forgette
3,149 Views

Hi Dan,

     Again, I apologies for the confusion.  The Datastore Provisioning Wizard is actually a feature of the RCU.  As I suggested in the related post, we've been getting this a lot lately, so I don't think it will be long before we clear things up.  😉

The RCU will provision and mount NFS datastores only if the ESX hosts' vmkernel interface and the NetApp controller all have ip addresses on the same network (in the same subnet).  This is a best practice that the RCU enforces to prevent performance issues from sneaking up later.  The same logic is used in the case of iSCSI.  The VSC and RCU are both fully supported NetApp products, so if you're having issues, please don't hesitate to open a support case.
Thanks for using our tools and for posting the questions (you're not the first person to run into this).
Cheers,
-Eric

View solution in original post

3 REPLIES 3

forgette
3,150 Views

Hi Dan,

     Again, I apologies for the confusion.  The Datastore Provisioning Wizard is actually a feature of the RCU.  As I suggested in the related post, we've been getting this a lot lately, so I don't think it will be long before we clear things up.  😉

The RCU will provision and mount NFS datastores only if the ESX hosts' vmkernel interface and the NetApp controller all have ip addresses on the same network (in the same subnet).  This is a best practice that the RCU enforces to prevent performance issues from sneaking up later.  The same logic is used in the case of iSCSI.  The VSC and RCU are both fully supported NetApp products, so if you're having issues, please don't hesitate to open a support case.
Thanks for using our tools and for posting the questions (you're not the first person to run into this).
Cheers,
-Eric

daniel_kaiser
3,149 Views

Hey Eric thanks for the advice.  First of all I totally forgot about RCU since we installed RCU and VSC to begin with but haven't looked at it since we weren't setting up any clones just yet.  We're still very green with the whole netapp and vmware integration.  Anyhow, i saw the Youtube video and saw that you can configure RCU to look at only certain ports on the netapp, so we went ahead and modified it so it will only look at the private network.  This change appeared to have fixed all the issues we were running into when provisioning new storage using the wizard and also running an update in the Netapp Virtual Storage Console tab.

Separate question, when I destroy a datastore it deletes the Volume but not the export.  Not sure if this is normal...

forgette
3,149 Views

The export not being cleaned up sounds like a bug, however I was not able to recreate it in my lab.  If this causes issues, you might want to open a case.  Also, be on the lookout for version 3.0 of the RCU.  It should be out next month.

Public