DL380-G3 / SmartArray 5i fails to restripe new drive

[oldtrlrnr]

This DL380-G3 has P29 system ROM 10-31-2003, and its embedded Smart Array 5i has firmware 2.56B, and accelerator enabled. It has 2 arrays, a RAID-1 mirror of 2x36GB and a RAID-5 array of 4 72GB drives, all 10Krpm. We got a batch of 146GB/15Krpm drives and want to swap in and restripe one drive at a time, then assign another logical drive to use the new capacity. We removed one drive and plugged in one of the new 146GB/15K drives. The ACU showed it did NOT start restriping it, in fact the new drive shows as unused, not in any array, and the slot the original drive was in shows with a red X and of course array degraded. I removed this drive and plugged in the original 72GB/10K drive, but ACU shows the same status, the original drive also showing as unused. I'm looking for a safe way to bring this new drive into the existing RAID-5 array and restripe without losing any data. Appreciate any suggestions.

[Srinivas Suddapalli]

Hi,

The best way before you make any changes to your array is to have a backup. Later get the server to updated PSP - Prolaint support pack and als update firmware using firmware disk as in your case the system rom and firmware of the controller are old.

Once you feel that the server is stabilized you can start with new drives.

When you connect a 15K drive in a 10k drives array, the drive would work at the least rpm drive.

Link: http://h20000.www2.hp.com/bizsupport/TechSupport/SoftwareIndex.jsp?lang=en&cc=us&prodNameId=3288130&prodTypeId=15351&prodSeriesId=316529&swLang=8&taskId=135&swEnvOID=1005#29214

Regards,
Srinivas S

[oldtrlrnr]

Of course I won't do anything before getting a full backup; in fact we've gotten that done.

Here's a little more background on this problem. My friend had two DL380-G3's. The one he'd had in production, hosting all his data, started crashing. He pulled all its drives and put them in the other one, in the corresponding hot-swap bays. The RAID-5 array may have been degraded ever since then, we don't know. After pulling one of the 72GB drives in the RAID-5 array and substituting a new 146GB/15Krpm drive, I checked status in the ACU and found the new drive was not in the array, which showed degraded status. One fact I think is significant is that the failed drive in the physical view of the array showed as SCSI port 1, ID 3, whereas the new drive showed as SCSI port 2, ID 3, even though it's the same bay. So we suspect that the port allocations differed between the two DL380's. Maybe the original and new DL380's SCSI backplanes were configured differently, simplex/duplex. Another question is whether the DL380-G3 even supports the 146GB/15Krpm drives. They are not shown on the drive list for its QuickSpecs (only 146GB/10Krpm), but that doc was written before the 146GB/15K drives came out.

We're still looking for a way to fix this degraded array and upgrade all its drives to 146GB without losing data. Our ideas include (1) checking the simplex/duplex configs of the backplanes on both servers and if different setting the new one to match the old one; and (2) installing a set of new blank drives, configuring the arrays, then powering down and substituting the existing set of drives with the real data.

Appreciate any suggestions.