Search This Blog

2012-10-17

Arcconf (Adaptec) RAID Commands

Finding a Dead Drive
In order to find a drive that has previously failed using adaptec's aarconf utility you would use the following options:

Command:

arcconf getlogs 1 DEAD tabular

Output :
Controllers found: 1

    ControllerLog
      controllerID .................................... 0
      type ............................................ 5
      time ............................................ 1349289004
      version ......................................... 1

       deadDriveEntry
         vendorID ........................................ XXX
         rtcSeconds ...................................... 52
         rtcMinutes ...................................... 6
         rtcHours ........................................ 13
         rtcDay .......................................... 24
         rtcMonth ........................................ 8
         rtcYear ......................................... 2012
         serialNumber .................................... XXX
         productID ....................................... XXX
         wwn ............................................. XXX
         failureReasonCode ............................... 2

It gives a good idea of drive failure rate on a controller and logs a history of previous failures and failure times.

Here is a table of what the 'failureReasonCode' parameter means:
0 or 0x00 Unknown failure
1 or 0x01 Device not ready
2 or 0x02 Selection timout
3 or 0x03 User marked the drive dead
4 or 0x04 Hardware error
5 or 0x05 Bad block
6 or 0x06 Retries failed
7 or 0x07 No Response from drive during discovery
8 or 0x08 Inquiry failed
9 or 0x09 Probe(Test Unit Ready/Start Stop Unit) failed
A or 0x0A Bus discovery failed 

Hot Spare

Ensure the drive is in a good and in ready state prior to assigning to a LD
arcconf setstate 1 device 0 22 rdy
Assign the drive as a hot spare to the logical drive labeled 1
arcconf setstate 1 device 0 22 hsp logicaldrive 1

To set a global hot spare that will hop in regardless of which array is degraded just do not specify a LD
arcconf setstate 1 device 0 22 HSP

Verify Disks

To verify disk 22 and to do so without a prompt use the following. It will scan the disk for defects and report back
arcconf task start 1 device 0 22 verify noprompt


Detailed Support Logs

arcconf savesupportarchive
Controllers found: 1

All the logs are saved in /var/log/Support

Command completed successfully.

tar -zcvf support.tgz /var/log/Support/

The support logs will give you
  • performance metrics
  • consolidated logs
  • history of events
  • progress of historical events
  • more detailed logs
  • number of commands
  • command completion time
  • typical block sizes for read/write
  • uncached vs cache io hits
  • recent command history

No comments:

Post a Comment