VNX

[beav_]

Is it possible to replace the first 4 drives in a VNX 5200?  They are 300GB and I would like to swap out for 1.2TB drives.  Actually, I plan to replace them for the entire enclosure. 

There are other DAE's attached for me to swing existing data to, but is that necessary or can I replace them  one-by-one if they are unbound?

Again though, I'm more concerned about how to handle the first four drives.  Is there a documented process for this?

Disclaimer, I am new to EMC storage and this system is being used for a development project.

Thank you!

[DELL-Josh Cr]

Hi,

Unfortunately, there is not a supported method for increasing the drive size.  

Let us know if you have any additional questions.

[beav_]

If one of the existing drives fails, can it be replaced with a larger drive?  If so, would there not be an opportunity to pre-fail the drives? 

On another note, can user data be stored on drives 0-3 or is it just reserved for the OE, etc?  If not, I may not be overly concerned with upgrading the capacity on the first four drives.

[DELL-Josh Cr]

Yes, if a drive fails it can be replaced with a larger drive but the extra space won’t be used.  The first 4 drives can have user data, only some of the drive is used by the system. How much will vary and also depend on the way the pool is configured. 

[beav_]

Thanks, I appreciate this info.  It helps me to understand my situation better.  Are 0-3 in their own dedicated pool/RAID group or can they be shared with the remainder of the drives in the main enclosure?

In any event, when 0-3 fail and are replaced with larger drives will that extra space be available for user data automatically?  

If so, can they be pre-failed manually?

[DELL-Josh Cr]

No, the drives can’t be expanded that way, you have to migrate data delete the pool and create a new pool with the new drives. The system drives are supposed to be separate from the storage but can be combined.

This is what goes on the system drives:

There are several private LUNS on these system drives including LUNs for:

- Operating system for both SPs in RAID1

- Persistent Storage Manager (PSM)

- Operating environment database in RAID 1 triple mirror

- The vault in RAID5.

This organization of the drives provides protection from two drive failures.

[beav_]

"you have to migrate data delete the pool and create a new pool with the new drives" 

This procedure would not work for my needs? Based on the statement it seems like I could migrate the pull from the system drives to four other temporary drives swap the larger drives and Bay 0-3 then migrate back?  

[DELL-Josh Cr · Accepted solution]

I am not sure it will migrate the system LUNs. It isn’t supported so it is at your own risk to try it.

[beav_]

I may check into it a bit but I don't want to break the system for sure. I've also wondered if a byte level copy outside the system drive for drive would work.

[Rainer_EMC]

no that wont work