Question about Replacing P4000 ESXi VSA OS DISK after vmware corrupted

[rocksoft7]

hi,

if we have p4000 vsa esxi on local datastore (local disk) and we attached external disk 5 x 2TB for vsa storage system..

all work fine until one day our vmware local disk corrupted so the VSA storage become down..

so at this point we know the vsa storage 5 x 2TB is still healthy since only vmware local disk is corrupted, so can we simply install new vmware server, create new vsa virtual machine, and attached the existing disk 5 x 2TB to the new vsa?

will the CMC detect it fines and allow us to access the old data inside 5 x 2TB disks array?

i did tried it, i even use same ip and mac address on the new vsa, but it seems the CMC have problem to recognized it still.. kindly refer attached image .. tqvm

any idea?

thanks

[a_o]

I'm pretty sure that it will not work the way you're thinking, as I've tried before and it did not work.

You have problems at two levels.

1, Your new VSA is not properly initialized.

2, CMC is confused about which VSA is which.

To solve these issues, just take advantage of the fact that SAN i/Q uses Network RAID, and you still have a copy of your data on the good VSA node:

First, at the SAN  i/Q console in VMware console, you should reset the new VSA that's using the existing  5 x 2TB disks, and give it a new different name.

Also, if possible, use a different IP address.(it's not critical, but it make things easier).
 Use the old MAC address,  as you would not need to Re-host on HP Poetic/license portal.

Then,  in CMC - on the new VSA  - expand to "Storage".

Select the "Disks" tab as shown in your screenshot and select a disk, RMC  and choose "Add Disk to RAID"

Do this for all the disks.

Go to the "RAID Setup" tab, RMC and select "Re-configure RAID".

Your RAID set should be normal after this. If not, just re-install the VSA again using the settings listed above.

Lastly, in CMC main menu, Choose "Find->Clear all Systems" .

Then,  add the new VSA to your existing Management Group and then to the cluster.

Your LUNs would then sync to the new node depending on which Network RAID setting you're using.

You could also just give support a call.

[rocksoft7]

hi,

1. thanks for your response, im my case this vsa is node on its own, the data dont sync to other storage or vsa..

but i still can follow your step to add the uninitialized disks to raid..and add the vsa node to cmc mgmt group..

in my case will it mean data loss ? add disk to raid will reformat it? leaving mgmt group will delete data on vsa external storage?

2. the 5 x 2TB disks still healthy and the data surely is there, so you are saying on single node vsa if we corrupted the vsa vmdk (vmware) , we will loss all storage data even the data is in external storage conencted to vsa?

3. if i send my broken local vmware hdd for data recovery and they managed to get the vmdk file,  and i install new vmware os and load this vmdk, will my vsa up again?

thanks

[a_o]

To your Question #1.

In my experience, yes it will mean data loss.

To your Question #2.
Not  exactly. Yes the data is still there. But it's  not in any meanigful state to a new VSA.

The VSA's config partition is where the RAID configuration and block to LUN 'mappings'  are stored. 

So when you create a new VSA, those 'mappings' are blank.

To your Question #3.

Yes, this would be a way to 'restore' your previous VSA.

You would just have to make sure that the datastores are  still mapped in the same manner as they were before in VMware.

That is , your new VMware install should mimic the previous one, WRT storage presentation.

It  seems that you've already done this.

However, if you go through the steps in your question #1, then you would be essentially overwriting the existing data with zeros. So your actual data would no longer be recoverable.