Need some help to understand RAID configuration 1

[nstefi1]

Hi guys,

I am working at a new company where the previous network admin left before I came and I can't figure out what is the situation with one of the servers.
One of the hard drives looks like somebody wanted to pull out and he changed his mind half way. The plastic is out on one side, but the hard disk is still in.
I checked with a software called "Intel Storage Console" and when I choose "Configure Host Drives" it shows:

Select Host Drive
No - Name - Status - Attr - Capacity - Type - belongs to
0 - RAID5 - ready - RW - 139945MB - RAID-5 - Array 0
when I press F4 for Drive Information, it shows one more time the same line with different title (Array/Logical Drive)
If I press F4 again, it shows the following:
RAID 1/Logical Drive(s)
No - Name - Status - Attr - Capacity - Type - belongs to
0 - PD_A000 - ok - RW - 69974MB - Disk - Master
1 - PD_A002 - ok - RW - 69974MB - Disk
2 - PD_A003 - ok - RW - 69974MB - Disk
3 - PD_A005 - ok - RW - 69974MB - Disk - Priv. Hot Fix D

Also HDD with ID2 has Grown Defects = 4, the others have 0.
Is this PD_A005 a hot spare disk? And what is Master? I thought data and parity should be spread out evenly on all drives in RAID5, except the spare drive.
If drive 2 is defective, how can I pull it out? Do I need to activate the hot spare first in the software and prepare the disk 2 to plug out, or just plug it out as it is and everything is automatic?
My RAID controller is SRCU31L.

Thanks for help.

 

[wahnula]

Look here:

http://www.intel.com/support/motherboards/server/srcu31l/sb/CS-016800.htm

Download the PDF, page 110 on, tells about the RAID 5+1 feature and how to replace a disk.

Tony

Users helping Users...

 

[nstefi1]

Thank you very much.

I didn't know there are so many drive types: physical, logical and host drives.
I thought logical drive means already what Windows shows as a disk, but I guess the host drive is what Windows shows, and each logical drive has a corresponding physical drive in my case.

I opened Intel Storage Console Pro and I save the information about the controller. I don't understand why I have so many messages in the section called "Screen Service messages / Async. events".
The same messages are repeating on certain days, but I don't know what is the cause. Is it because of a server restart? Or because somebody unplugged or replaced a hard disk? I post here a part of the log.

Thanks


Mon Sep 15 11:20:37 2008: RAID-5 Host Drive 0 installed (ready)
Mon Sep 15 11:20:35 2008: SCSI-A ID:5 LUN:0 -- SEAGATE ST373207LC 160MB/s
Mon Sep 15 11:20:35 2008: SCSI-A ID:3 LUN:0 -- SEAGATE ST373207LC 160MB/s
Mon Sep 15 11:20:34 2008: SCSI-A ID:2 LUN:0 -- SEAGATE ST373207LC 160MB/s
Mon Sep 15 11:20:34 2008: SCSI-A ID:0 LUN:0 -- SEAGATE ST373207LC 160MB/s
0:0:0:34: Realtime: Mon Sep 15 11:20:28 2008
0:0:0:17: ECC support enabled
0:0:0:6: SCSI-A ID:6 LUN:0 -- ESG-SHV SCA HSBP M16
0:0:0:4: SN 00F98B2C - Firmware-FW-Version 2.32.02-R00F - Nov 7 2001
0:0:0:4: SRCU31L - HWL0 - 16 MB SDRAM/ECC - 2048kB Flash-RAM
Thu Sep 04 08:28:56 2008: RAID-5 Host Drive 0 installed (ready)
Thu Sep 04 08:28:54 2008: SCSI-A ID:5 LUN:0 -- SEAGATE ST373207LC 160MB/s
Thu Sep 04 08:28:54 2008: SCSI-A ID:3 LUN:0 -- SEAGATE ST373207LC 160MB/s
Thu Sep 04 08:28:53 2008: SCSI-A ID:2 LUN:0 -- SEAGATE ST373207LC 160MB/s
Thu Sep 04 08:28:53 2008: SCSI-A ID:0 LUN:0 -- SEAGATE ST373207LC 160MB/s
0:0:0:34: Realtime: Thu Sep 04 08:28:47 2008
0:0:0:17: ECC support enabled
0:0:0:6: SCSI-A ID:6 LUN:0 -- ESG-SHV SCA HSBP M16
0:0:0:4: SN 00F98B2C - Firmware-FW-Version 2.32.02-R00F - Nov 7 2001
0:0:0:4: SRCU31L - HWL0 - 16 MB SDRAM/ECC - 2048kB Flash-RAM
Thu Aug 21 06:02:30 2008: RAID-5 Host Drive 0 installed (ready)
Thu Aug 21 06:02:28 2008: SCSI-A ID:5 LUN:0 -- SEAGATE ST373207LC 160MB/s
Thu Aug 21 06:02:28 2008: SCSI-A ID:3 LUN:0 -- SEAGATE ST373207LC 160MB/s
Thu Aug 21 06:02:27 2008: SCSI-A ID:2 LUN:0 -- SEAGATE ST373207LC 160MB/s
Thu Aug 21 06:02:27 2008: SCSI-A ID:0 LUN:0 -- SEAGATE ST373207LC 160MB/s
0:0:0:34: Realtime: Thu Aug 21 06:02:21 2008
0:0:0:17: ECC support enabled
0:0:0:6: SCSI-A ID:6 LUN:0 -- ESG-SHV SCA HSBP M16
0:0:0:4: SN 00F98B2C - Firmware-FW-Version 2.32.02-R00F - Nov 7 2001
0:0:0:4: SRCU31L - HWL0 - 16 MB SDRAM/ECC - 2048kB Flash-RAM
Fri Aug 08 06:02:24 2008: RAID-5 Host Drive 0 installed (ready)
Fri Aug 08 06:02:23 2008: SCSI-A ID:5 LUN:0 -- SEAGATE ST373207LC 160MB/s
Fri Aug 08 06:02:22 2008: SCSI-A ID:3 LUN:0 -- SEAGATE ST373207LC 160MB/s
Fri Aug 08 06:02:22 2008: SCSI-A ID:2 LUN:0 -- SEAGATE ST373207LC 160MB/s
Fri Aug 08 06:02:21 2008: SCSI-A ID:0 LUN:0 -- SEAGATE ST373207LC 160MB/s
0:0:0:33: Realtime: Fri Aug 08 06:02:15 2008
0:0:0:17: ECC support enabled
0:0:0:6: SCSI-A ID:6 LUN:0 -- ESG-SHV SCA HSBP M16
0:0:0:4: SN 00F98B2C - Firmware-FW-Version 2.32.02-R00F - Nov 7 2001
0:0:0:4: SRCU31L - HWL0 - 16 MB SDRAM/ECC - 2048kB Flash-RAM

 

[wahnula]

To me it just looks like a normal log, no events started or stopped, no warnings, no rebuilding. Might be coincident with server restarts, might not, but unless I'm missing something, this is strictly informational, no warnings that I can detect.

Here's a snippet of my 3Ware log that had a problem:

E=0202 I=008EE65C T=22:28:09     : Data ECC error (host)

       ata task file written out : cd dh ch cl sn sc ft

                                 : C8 E3 AE 7F BF 20 00

         ata task file read back : st dh ch cl sn sc er

                                 : 51 E3 AE 7F D9 20 40

E=0202 I=008EE65C T=22:28:09 P=1 : Repair LBA 0x3AE7FD8...OK

E=0202 I=008EE65C T=22:28:09 P=1 : Repair LBA 0x3AE7FD9...OK

E=0202 I=008EE65C T=22:28:09 P=1 : Repair LBA 0x3AE7FDA...OK

Send AEN (code, time): 0x23, 09/17/2008 22:28:09

Sector repair completed

(EC:0x23, SK=0x01, ASC=0x11, ASCQ=0x00, SEV=02, Type=0x71)

port=1, LBA=0x3AE7FD9

E=0202 I=008EE65C T=22:28:09 P=1 : Prepare for command retry

E=0202 I=008EE65C T=22:28:09 U=0 : Retrying command in degraded read mode

-- Host message: 0.0.0  166) +Warn+ AEN: SECTOR_REPAIR (port=1, LBA=0x3AE7FD9)

Saving PRINTLOG, time=4278721 ...  
### END Firmware Print Log

As you can see, it's clear when there's a problem, even with a convoluted log. And I would not worry about the (4) Grown Defects, that's a tiny amount. And for future reference, most RAID controllers are fully automatic, when you remove a drive and replace it with a freshly-formatted new drive, it takes over from there. RAID 5+1 is even more secure.

Tony

Users helping Users...