ESX Ranger, ESX, Netapp, NFS

If you have ESX running on Netapp, you sure as hell probably installed the Netapp ESX tools. If you didnt then you should. A couple things that the tools do is change NFS settings within ESX.

I ran into an issue with ESX Ranger failing on backups from Netapp NFS to a Netapp CIFS share. The error was:

The Backup file /vmfs/volumes/NFS_TIER_BLAH/Server1//Server1-flat.vmdk that was transfered appears to be Invalid! Transfered Size: 112546265463, Expected Size: 94667312308.

The solution seems to be changing the NFS timeout values within ESX by:

Changing the value of NFS.HeartbeatFrequency to 12
Changing the value of NFS.HeartbeatMaxFailures to 10

This will work for most people if they havent already installed the Netapp tools and therefore already have these settings. This didnt fix my problem, but it may yours.

My solution was to Use NFS instead of CIFS for the target. It works thus far.

Sean

Some Basic Netapp Tips 1

Here are some tips that I share with clients. They are not rules of thumb, simply ways of doing things. For most of my clients I share a VM Tips file, which I update constantly and a storage tips file, in this case Netapp.

After creating volumes log into the filer and turn on volume autosize with the trigger set to volume

After creating volumes log into the filer and set the snap autodelete to on

Set the snapreserve % to 0 when creating volumes that will house LUNs. Just create the volume big enough to handle a LUN plus the snaps manually.

Make sure that when you grow a volume that you login to the filer and reconfigure volume autosize. This can be done by typing vol autosize reset; followed by vol autosize on

Always set the default security to restricted for iSCSI and add initiators as needed

Continued.....

Extreme Networks ESX Port Grouping

By default if you create a port group on an Extreme switch it is layer 2. If you then create a port group in ESX or a vswitch and choose IP hash for load balancing, it wont work properly. You need to change the port grouping on the switch to L3_L4.

The same applies to a Netapp dynamic VIF as Netapp uses IP load balancing as well.

Netapp and ESX Network Setup

Here is a quick snapsnot of a setup I have done in the past with ESX and Netapp

Netapp Volume Creation

I have decided to make a quick reminder of things that I usually do after creating a Netapp volume:

vol autosize (vol) on - This will autogrow the volume once a threshold is reached
snap autodelete (vol) on - This will auotmatically delete snapshots when a threshold is reached
sis on (vol) - This turns on single instance storage. You can check the status of SIS by running a df -s
Change the security on the NFS mount for NFS volumes to allow for the IPs of the servers that will use the volume.

Normally this is enough. I will have more information if you are creating a volume that is going to be SnapMirrored.

Netapp SnapMirror with SMVI

If you are using SnapManager for Virtual Infrastructure with SnapMirror, you may have seen this error:

[ERROR - SnapMirror update from source snapshot null to destination location DESTFILER:dest_vol failed]

The reason for this is that I have edited the snapmirror.conf file in order to force SnapMirror to use a specific NIC port.

When you do this SMVI cannot initiate SnapMirror because it expects a different filername in the SnapMirror.conf. Since I changed the name to point to the other interface, you must do the following to the snapmirror.conf in order for SMVI to now work:

(sourcefilername) = multi (sourcefilerIP,TargetFilerIP)