topic Re: Proliant ML370 RAID Rebuild Question in ProLiant Servers (ML,DL,SL)

Proliant ML370 RAID Rebuild Question

Chris Thornberry — Sat, 04 Mar 2006 10:44:30 GMT

I am working on a Proliant ML370 Server with a Smart Array Controller 6400. The controller has two logical disks configured as follows:

Parallel SCSI Array A (69452MB, RAID 1+0)
72.8GB Drive 1:4
72.8GB Drive 1:5
Parallel SCSI Array B (420027MB, RAID 5)
146.8GB Drive 2:0
146.8GB Drive 2:1
146.8GB Drive 2:2
146.8GB Drive 2:3

Ok, so yesterday I walked into the server room and physical drive 2:1 had failed. I contact HP support and they sent out a replacement drive. Well, I also noticed their battery back had issues with a failed battery so we swapped out the battery backup unit last night when we could reboot the server.

Well, upon rebooting of the server the adapter determined the drive in port 2:1 was good and needed to be rebuilt. SO, the controller rebuilt the array on the "apparent bad" disk.

Ok, that's not the problem though ... the problem is that it currently has a message:

"Background parity initialization is currently queued or in progress on logical drive 2 ( RAID 5 in array B). If background parity initialization is queued, it will start when I/O is performed on the drive. When background parity initialization completes, the performance of the logical drive will improve."

By the way, we do not have a parity drive configured on this array.

Now it has said this for close to 10-11 hours now. I am used to other manufacturers and their RAID utility and you are able to see a % complete and I cannot find anywhere that it has any sort of progress.

They currently have 117GB of used space on the array. I have heard conflicting reports of the time it takes to finish a rebuild ranging from 1min3secs/GB of used space to 15mins/GB of used space.

Can anyone shed any light on this for me?

Thanks in advance ...

Re: Proliant ML370 RAID Rebuild Question

Igor Karasik — Sat, 04 Mar 2006 11:17:31 GMT

Chris,
Array rebuild may take 15 min for 1 Gb of data (and even more - depends on rebuild priority). Don't worrry about it.
You can check Array rebuild progress in System management homepage->SmartArray 6400 controller link (system management homepage is part of Proliant support pack)

"By the way, we do not have a parity drive configured on this array"
But you have RAID-5 , right ?

BTW, battery pack message CAN be related to outdated SmartArray 6400 firmware
http://h18000.www1.hp.com/support/files/server/us/revision/7717.html
"Updated the battery charging firmware to reduce the frequency and duration of battery charging cycles. This helps prevent overcharging the battery, which can contribute to premature battery failures. This also reduces the time required for a battery recharge cycle to complete. Updated the battery firmware to improve the sensing and reporting of battery health. This helps prevent battery status messages from alternating or indicating false status. "
So, update your SmartArray 6400 firmware to 2.58 version (do it AFTER rebuild completed)

Re: Proliant ML370 RAID Rebuild Question

Chris Thornberry — Sat, 04 Mar 2006 11:44:06 GMT

Igor, thanks for the info. The unfortunate part is that I am unable to access the HP Insight Manager Homepage. I bring up the web page and it simply sits there for a long time and never times out or displays anything.

I have a case open with support and they are analyzing my ADU output. The part that is sort of "scary" to me is the drive was identified as bad before and now my array is being rebuilt on a drive that was bad before. I have the replacement drive on-hand but am leery (sp?) of replacing the old drive while the array is rebuilding.

I did the math and approximately 120GB x 15min/GB = 30 hours ... so we'll see sometime tomorrow what finally happens.

But any ideas on the Insight manager not working?

Re: Proliant ML370 RAID Rebuild Question

Igor Karasik — Sat, 04 Mar 2006 12:39:51 GMT

Chris,
System management homepage (SMH) (you write "HP Insight Manager Homepage", but Insight Manager is completely different thing, I think you mean system management homepage)) timeouts occur because one of the applications in the SMH is not responding.
Common reasons for such behaviour:
-version control agent (VCA) is not configured to point to version control repository server
-SNMP misconfiguration
If you use windows 2003 most likely you don't have "localhost" entry in the security properties of SNMP service under "Accept SNMP packets from these hosts"
See this advisory as well
http://h20000.www2.hp.com/bizsupport/TechSupport/Document.jsp?objectID=PSD_EM021108_CW01

Regarding "bad" drive that going to be "good" after reboot:
Rarely I saw situation that take out/put back "bad" drive "convert" it to be "good". It maybe outdated SmartArray controller or SCSI drive firmware problem as well.

Re: Proliant ML370 RAID Rebuild Question

Chris Saindon — Mon, 06 Mar 2006 11:46:13 GMT

Hello Everyone,

As for the homepage, double check that SNMP is configured correctly, make sure you're using the current Support pack 7.40, and also verify that Java is installed (min v1.40). When installing the support pack, there are 2 things that must be configured under "All configurable components", HP System Management Homepage, and HP Management Agents.

Proliant Support Pack 7.40 (Windows 2000)
http://h18023.www1.hp.com/support/files/server/us/download/23836.html

Proliant Support Pack 7.40 (Windows 2003)
http://h18000.www1.hp.com/support/files/server/us/download/23837.html

With this you should be able to get in.

Regards
Chris