RAID0 to RAID1

[Binand Sethumadhavan]

Hi,

I have a DL360G3, running RHEL3. Now I want to insert another disk into the box and move over to RAID1. The disk is already in, now is there a way I can change the RAID level without having to reinstall? I tried (from within hpacucli):

controller slot=0 array A add drives=1:1
controller slot=0 logicaldrive 1 modify raid=1

But it didn't work :( Gave me:

Error: This operation is not supported with the current configuration. Use
the "show" command on devices to show additional details about the
configuration.

Now I can see it expanding.

controller slot=0 logicaldrive 1 show

Status: Expanding 9% complete

How do I stop this expansion and make it into a RAID1 group without losing data or having to reinstall?

Thanks,

Binand

[Vijay_44 · Accepted solution]

At the HW level (independent of the OS)expansion would be the first step. Once you EXPAND the array you will get unused space under the array, you can then MIGRATE the logical drive from RAID 0 to RAID 1, using that newly acquired unused space.

[Binand Sethumadhavan]

Hi,

Thanks for that. Worked like a charm. Though, with 2x72GB disks, it took about 10 hours (I didn't time it, but it ran overnight) to migrate from RAID0 to RAID1 - I take it this is normal?

Now I have done the same on two other servers (all three are identical). On one server I see the expansion progressing nicely:

Status: Expanding 16% complete

But on the other server, it doesn't progress at all:

Status: Expanding 0% complete

It has been sitting there even though I started both of them almost simultaneously.

I am getting worried here. Is it going to crash my server? Should I start taking an emergency backup? The only difference so far is that while the first two servers were idle, the third one is running a reasonably busy mail relay (kicks self for doing this on a production system).

Thanks,

Binand

[Binand Sethumadhavan]

Some more news on this.

I ran this procedure on many more servers. Here is what I see now:

carbuncle: Status: Expanding 81% complete
schwartz: Status: Expanding 57% complete
stallman: Status: Expanding 54% complete
thomson: Status: Expanding 53% complete
torvalds: Status: Expanding 52% complete
vixie: Status: Expanding 52% complete
volkerding: Status: Expanding 53% complete
wall: Status: Expanding 1% complete

This server named "wall" is stuck at 1% - even though the others were started much later. Is it because the server is busy and hence the disk state does not remain consistent? Or is there something else with this server that I need to look into?

Thanks,

Binand

[Vijay_44]

As you have already found out, the time take to re strip (expand, extend, rebuild, migrate) is inconsistent and server activity has a direct impact on how long the process would take. I think in the ACU there is a setting for Expand priority, I would try to increase it and see if the process is faster. For the "wall" I would give it some time to see if the progress improves. If not try up date the FW on the HDD and the array controller.


FW can be downloaded from

http://h18000.www1.hp.com/support/files/storage/us/index.html


Always a good idea to have a backup before any array modifications

[Mitch Sprague]

Thanks to all. This was exactly what I was looking for and it works like a charm.