vSphere 5 Datastores

Hi,

We've just upgraded to vSphere 5.0.1 and I'm going to blow away our datastores to reformat as VMFS 5. I've moved all the VMs off the datastores but there is a folder called ".vsphere-ha" still left on a couple . Can i safely delete this or is there something else we need to do.

Cheers
kswan_expertAsked:
Who is Participating?

Improve company productivity with a Business Account.Sign Up

x
 
coolsport00Connect With a Mentor Commented:
For removing datastores, yes, you don't need to undo HA as mentioned above. I personally do when I'm doing Host maintenance, but you're only removing/re-adding datastores. You can delete and it should be fine. I have yet to have any issues with deleting a datastore in the 5yrs I've been using VMware.

Yes, once you remove the datastore, however you choose to, then yes, just remove the LUN on your SAN then recreate. Attach to your ESXi Hosts on your SAN, then rescan your HBAs on each Host, then re-add the datastore thru the 'Add Storage' wizard (Config tab > Storage > Add Storage link).

Regards.
~coolsport00
0
 
coolsport00Commented:
There should be a file within that folder that is used for HA; specifically, HA Datastore Heartbeating. You can disable HA in your Cluster then re-enable it..that should remove the folder/file. Once you're done upgrading your DSs, re-enable HA which will then auto recreate the files on the DSs. HA uses an algorithm based on a few factors to use for DS Heartbeating, which is a 2nd 'tier', if you will, used in determining if a host is isolated or not.

~coolsport00
0
 
kswan_expertAuthor Commented:
Cheers coolsport.

Lucky i asked this as i'd never thoughg abut disabling HA

So disable HA on the cluster, then folders should disappear and i right click on datastore,  then do i unmount or delete?  Then i go to the SAN,  delete the storage and recreate? I'm going to recreate larger datasores an re-present these.


Cheers
0
Keep up with what's happening at Experts Exchange!

Sign up to receive Decoded, a new monthly digest with product updates, feature release info, continuing education opportunities, and more.

 
Andrew Hancock (VMware vExpert / EE MVE^2)VMware and Virtualization ConsultantCommented:
You can safely delete, and it does not matter about Disabling/Enabling HA, VMware HA will re-create it!

Technically, you should delete it, and then re-scan datastore, and then Mask Out, re-scan, then remove from your SAN.

(make sure you haved masked out on ALL ESXi servers, there is a nasty bug which still exists which sends the iSCSI initiator into a "dive", causing a hang for several minutes on ESXi 5.x (yes all versions!)

Create new LUN, remove LUN Mask, Add to ESXi (which will partiton and format as VMFS-5).

Ensure that it's ACL/Permissions are allocated to ALL ESXi hosts.

Re-scan and it should be then presented to ALL ESXi hosts.
Masking a LUN from ESX and ESXi using the MASK_PATH plug-in
0
 
kswan_expertAuthor Commented:
Cheers hancocka.

I'm getting a bit confused now, but that's just me (:

Do i want to an unmount or delete, whats the difference?? Also, how\why do I mask out?

I was going to go to HOME > INVENTORY > DATASTORES ... ensure there were no VMs or files on the datastore, right click and DELETE. Then go to HOME > INVENTORY > HOSTS and CLUSTERs and rescan each host. Then to SAN and delete LUN, recreate and re-present as normal. Sound safe and sound?

I don't mind turining off HA while i do this so i might if it is "safer"

Cheers
0
 
Andrew Hancock (VMware vExpert / EE MVE^2)Connect With a Mentor VMware and Virtualization ConsultantCommented:
the article for mask is in my post.

you mask, because there is a potential issue, when you delete the datastore from your first ESXi host, it goes "missing" on the other hosts, (which it would, because you have got no time to do anything with them) and this can cause a potential issue where the ESXi server hangs for minutes or hours.

Technicall, you should always mask first.

You can, work back from the SAN, and mask on the SAN first, this would ensure the datastore disappears from ALL hosts at the same time!

Your idea is sound, other than a bug which exists, which "may/can" cause the iSCSI initiator to try constantly for the missing LUN/datastore on other hosts, under certain conditions.
0
 
kswan_expertAuthor Commented:
Thanks guys, appreciate it.
0
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.