MSA1000 replacing failed disk

[Jim Geier_1]

I help manage an MSA1000 that is serving LUNs to a few OpenVMS systems. I have managed many HSG80 storage systems and more recently EVA8000 systems, and found the MSA1000 to be a little different. The systems and the MSA1000 are more than 300 miles away. I configured the MSA1000 while on site, but now my contact with the MSA1000 is via the MSA_UTIL program provided with OpenVMS. MSA_UTIL communicates with the MSA1000 over the fibre channel, but has limited commands compared to using the special serial cable and connecting directly to the controllers.

The MSA1000 appears to be running firmware version 4.48 (that's what I see when I issue the SHOW THIS or SHOW OTHER commands).

All LUNs configured on the MSA1000 were set up as RAID-6 or ADG. Recently, the display on one of the controllers reported: DRIVE FAILURE BOX #1, BAY 10. Two LUNs are on box #1, bay 10, and those LUNs each are spread out over 5 146 GB disks.

It does not appear that the spare disks automatically replaced this disk as I was expecting. Also, if I issue the command SHOW UNITS, it appears that there are active partitions on the supposedly failed disk. Prior to the failure, there were two partitions in use on that disk, partitions 0 and 1. Now, the SHOW UNITS command shows the LUN that was using partition 0 is using partition 255, and the LUN that was using partition 1 is using partition 0.

I highly suspect that this disk should be replaced because the amber light is flashing which almost always indicates a problem. What is the procedure for replacing a disk in the MSA1000 where two RAID-6 LUNs are using partitions on the disk? Or, where can I find that procedure?

[IBaltay]

Hi,
maybe this thread can help:
http://forums12.itrc.hp.com/service/forums/questionanswer.do?admit=109447627+1218083926952+28353475&threadId=1052061