Identifying failed hard drive in RAID

Question

I have a server which is running customized version of Debian. It is attached a sun storage raid. It has very limited tools and installing new tools is not allowed. :(

This message I see in dmesg:

end_request: critical target error, dev sda, sector 556782970
sd 0:0:0:0: [sda] Unhandled sense code
sd 0:0:0:0: [sda]  Result: hostbyte=0x10 driverbyte=0x08
sd 0:0:0:0: [sda]  Sense Key : 0x4 [current] 
sd 0:0:0:0: [sda]  ASC=0x44 ASCQ=0x0
sd 0:0:0:0: [sda] CDB: cdb[0]=0x28: 28 00 21 2f d5 7a 00 00 08 00
end_request: critical target error, dev sda, sector 556782970

So it seems there is badblock in sector 556782970, but I don't know which hard drive it belongs to so that I can get it replaced.

• arcconf getlogs command does not reveal any issues with drives.

• I did arcconf task start 1 logicaldrive 0 verify_fix but this didn't help.

• I did an e2fsck check. It attempts to fix some inodes but above issue remained. (I assume since its physical issue).

more info: http://pastebin.com/cJ2bUywj

Tools not available :( smartctl badblocks

Answer 1

If you actually use arcconf you can see physical drive status like this:

arcconf getconfig 1 PD

Look for drives with Failed state to identify drives that have been marked as failed. For your reference the output would look something like this:

  Device #6
     Device is a Hard drive
     State                              : Failed
     Block Size                         : Unknown
     Supported                          : Yes
     Reported Channel,Device(T:L)       : 0,15(15:0)
     Vendor                             : *MISSING*
     Model                              : 
     Firmware                           : 
     Total Size                         : 0 MB
     Write Cache                        : Unknown
     FRU                                : None
     S.M.A.R.T.                         : No
     S.M.A.R.T. warnings                : 0
     SSD                                : No
     MaxCache Capable                   : No
     MaxCache Assigned                  : No

Answer 2

Based on the end_request: critical target error, dev sda, sector 556782970 line, I assume that /dev/sda is the trouble child. You can find the serial number of that device using:

ls -l /dev/disk/by-id | grep "sda$"

Answer 3

While there might be an answer that gets you the info in the OS, odds are the controller firmware can tell you if you access it during boot.

Another thing I noticed in your cut and paste is that your battery says "failed". I wonder if the stripes are reporting failure because the Write Cache battery is toast? Though it reporting and actual bad inode would probably mean you have two problems.

Though my experience with Adaptec Controllers typically when the battery dies, it disables write caching as a preemptive measure.