Which disk in the mirror is having the problems?

Troy_Johnson@Dell.com Troy_Johnson at Dell.com
Thu Mar 17 14:08:05 CST 2005


Hi Jonas,

Use the Dellmgr-5.32-0.i386.rpm in the perc-apps tgz. http://support.dell.com/support/downloads/format.aspx?c=us&l=en&s=gen&SystemID=PWE_PNT_P3C_2850&os=LE30&osl=en&deviceid=1370&devlib=35&category=0&releaseid=R81687 

once installed just type 

dellmgr 

it will open a ncurses console. 

Go to objects/physical drive and highlight the suspect harddrive. Hit F2 and it will display any media errors or other errors (scsi bus errors) for that drive.

Troy
 
 

  _____  

From: linux-poweredge-admin-Lists On Behalf Of Jonas Blåberg
Sent: Tuesday, January 18, 2005 9:13 AM
To: linux-poweredge-Lists
Subject: RE: Which disk in the mirror is having the problems?



Hello again! 

I found some more info: The internal controller is a PERC3/Di (internal to PE2650). 

The extra controller says the following when booting: 

Jan 18 04:33:49 seprol03 kernel: scsi1 : LSI Logic MegaRAID 1.92 254 commands 16 targs 5 chans 7 luns 
Jan 18 04:33:49 seprol03 kernel: blk: queue f6f4ea18, I/O limit 4095Mb (mask 0xffffffff) 
Jan 18 04:33:49 seprol03 kernel: scsi1: scanning channel 0 for devices. 
Jan 18 04:33:49 seprol03 kernel:   Vendor: DELL      Model: PERC 3/DC         Rev: 1.92 
Jan 18 04:33:49 seprol03 kernel:   Type:   Processor                          ANSI SCSI revision: 02 
Jan 18 04:33:49 seprol03 kernel: blk: queue f6afdc18, I/O limit 4095Mb (mask 0xffffffff) 
Jan 18 04:33:49 seprol03 kernel:   Vendor: DELL      Model: PV22XS            Rev: E.14 
Jan 18 04:33:49 seprol03 kernel:   Type:   Processor                          ANSI SCSI revision: 03 
Jan 18 04:33:49 seprol03 kernel: blk: queue f6a98c18, I/O limit 4095Mb (mask 0xffffffff) 
Jan 18 04:33:49 seprol03 kernel: scsi1: scanning channel 1 for devices. 
Jan 18 04:33:49 seprol03 kernel: scsi1: scanning virtual channel 1 for logical drives. 
Jan 18 04:33:49 seprol03 kernel:   Vendor: MegaRAID  Model: LD 0 RAID1   69G  Rev: 1.92 
Jan 18 04:33:49 seprol03 kernel:   Type:   Direct-Access                      ANSI SCSI revision: 02 
Jan 18 04:33:49 seprol03 kernel: blk: queue f6a31c18, I/O limit 4095Mb (mask 0xffffffff) 
Jan 18 04:33:49 seprol03 kernel: scsi1: scanning virtual channel 2 for logical drives. 
Jan 18 04:33:50 seprol03 kernel: scsi1: scanning virtual channel 3 for logical drives. 
Jan 18 04:33:50 seprol03 kernel: scsi1: scanning virtual channel 4 for logical drives. 
Jan 18 04:33:50 seprol03 kernel: Attached scsi disk sdb at scsi1, channel 2, id 0, lun 0 

Can anyone tell me how to access this controller (PERC3/DC) now to check the defects lists of the disks?!?! 

/blåberg 


_____________________________________________ 
From:   Jonas Blåberg  
Sent:   den 18 januari 2005 15:53 
To:     'linux-poweredge at dell.com' 
Subject:        Which disk in the mirror is having the problems? 

Hello! 

I got some SCSI errors in the messages log on a RedHat server: 

Jan 18 04:29:56 seprol03 kernel: SCSI disk error : host 1 channel 2 id 0 lun 0 return code = 25040005 
Jan 18 04:29:56 seprol03 kernel:  I/O error: dev 08:12, sector 16 
Jan 18 04:29:56 seprol03 kernel: SCSI disk error : host 1 channel 2 id 0 lun 0 return code = 25040005 
Jan 18 04:29:56 seprol03 kernel:  I/O error: dev 08:18, sector 32780 

There are several other SCSI errors in the log but only devices 08:12 and 08:18 are shown, and 08:18 are many. 


So, does "dev 08:12" mean /dev/sda12 and "dev 08:18" mean /dev/sdb2? 

brw-rw----    1 root     disk       8,  12 Mar 19  2002 /dev/sda12 
brw-rw----    1 root     disk       8,  18 Mar 19  2002 /dev/sdb2 

/dev/sda12 is currently not used. /dev/sdb2 is used very much and it is a raw device, and it is the processes using the raw device which is having severe problems.


Both /dev/sda and /dev/sdb are mirrored devices (RAID 1). There are two different RAID controllers in the machine. 

The RAID controller where /dev/sda is located is shown in the admin interface "afacli". (I have opened the afa0 controller and run disk show defects 0 and 1 and I get nothing in the "grown defects" list - is there more to check for the disks?).

However I don't know how to access the other controller where /dev/sdb is located. Is there another admin interface available? I think something about LSI is said when the machine is booting.


Is there another way to check which disk in the mirror is causing the problems? 

Regards 
/blåberg 



-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.us.dell.com/pipermail/linux-poweredge/attachments/20050317/88440195/attachment.htm


More information about the Linux-PowerEdge mailing list