Expand RAID array (Replicated Volume)

I noticed some small print on support page:
http://mycusthelp.info/LSI/_cs/Answe...RJQLIIVYEULUZO[RHRQMZ&inc=8439&caller=1&txtCriteria=expand

NOTE: Capacity expansion is only possible on RAID 0, 1, 5 and 6. Please keep in mind, there should only be "1 virtual drive" on the array. If you have more than 1 virtual drive, then you cannot enlarge your array. You will have to create a backup, delete the array, create a new RAID and restore your backup to the new virtual drive.

Is this noticed anywhere else? (I didn't find it!).

I have the exactly same "no go" configuration with RAID5 array and 2 VDs: 20GB for OpenE OS and 5TB datastore which I try to expand to 6TB.

So, I have to:
- create "all new" system with separated System HDD and larger virtual drive.
- prepare it for replication from Active node
- replicate (sync) all data
- switch active/pasive nodes
- repeat the exercise on 2nd node

Any better idea?
Is there any utility to "migrate" DSS7 system VD to separate HDD?

This is the first time I have seen this. If you try to expand the raid by using the megaraid storage manager, are you prevented from doing so?

It's first time for me too!
As I explained before, the Drive Group expansion is only possible with single Virtual Disk.

So I planning to clone SSD7 system VD to phisical HDD, try to boot from new location and then remove this VD from array.
When only datastore VD will be in array, the expansion should be possible (it was approved by LSI Support, but we'll see).

(I use 6Gb/s RS25AB080 ROC, maybe this is solved on 12Gb/s controllers?)


Preliminary report - it goes somehow:

Cloning VD with SSD7 system was successfuly made sector by sector (Clonezilla), booting from new separate HDD was the same but with new activation.

Next step was removing "old" VD from array, I could not directly delete SYS VD - some complain about system disk. I erased content of this VD as first step and then deleted it.

Expanding array with new disk is allowed only in Bios Megaraid Utility.

The new array is under reconstruction (it takes about 35 hours!?!) and SSD Cache + WB Policy is automatically disabled.

...will be continued...

after 24h:
"E pur si muove"
but it goes annoing slow!
I'm just thinking, currently on a philosophical level, how to migrate:
- present MegaRaid VirtualDisk (R5 array of 8x 1TB disks), presented in DSS as Vg00
- to new (R5 array of 4x 3TB disks), will be also presented as Vg00 at node restart,
- and keep metadata so consistent, that replicated Logical Volunes will synchronise automatically - some kind sector-by-sector cloning.

And cluster must be online all the time.

...will be continued...