hi friends. In interview i was asked a question that. If a server is located at some remote place and in a server one of the hard disk had failed. Then how do we find which hard disk had failed from the several hard disk in the server remotely. we need to hav certain command for this in linux. in IBM AIX there is certain command but in linux i dont't know. does any 1 know this than plz reply.
Explore your hobbies and interests. Go to http://in.promos.yahoo.com/groups/
On Wednesday 02 Jul 2008, ravi kuamar wrote:
Then how do we find which hard disk had failed from the several hard disk in the server remotely. we need to hav certain command for this in linux.
Individual disks or hardware RAID?
in IBM AIX there is certain command
what is the command in AIX?
-- Arun Khan
On Wed, Jul 2, 2008 at 3:29 PM, Arun Khan knura@yahoo.com wrote:
in IBM AIX there is certain command
what is the command in AIX?
errpt -a or diag (or smitty diag)
Regards, NMK.
On Wed, Jul 2, 2008 at 3:29 PM, Arun Khan knura@yahoo.com wrote:
On Wednesday 02 Jul 2008, ravi kuamar wrote:
Then how do we find which hard disk had failed from the several hard disk in the server remotely. we need to hav certain command for this in linux.
Individual disks or hardware RAID?
If the disk is a member of a raid array, you can do a cat /proc/mdstat to know its health.
If its an individual disk, I don't know of any direct, linux native way of finding it. Your hardware vendor might provide some daignostic software that can be run without rebooting the system.
Regards, NMK.
On Wednesday 02 Jul 2008 19:35, Nadeem M. Khan wrote:
On Wed, Jul 2, 2008 at 3:29 PM, Arun Khan knura@yahoo.com wrote:
On Wednesday 02 Jul 2008, ravi kuamar wrote:
Then how do we find which hard disk had failed from the several hard disk in the server remotely. we need to hav certain command for this in linux.
Individual disks or hardware RAID?
If the disk is a member of a raid array, you can do a cat /proc/mdstat to know its health.
If its an individual disk, I don't know of any direct, linux native way of finding it.
grep syslog and dmesg.
On Wed, Jul 2, 2008 at 6:16 PM, jtd jtd@mtnl.net.in wrote:
grep syslog and dmesg.
True. This was a typical case of me thinking of complicated solutions and ignoring the most basic and simple one. :-) Does an amber light on the disk always trigger an alert in the syslog?
Regards, NMK.
On Wednesday 02 Jul 2008, Nadeem M. Khan wrote:
On Wed, Jul 2, 2008 at 3:29 PM, Arun Khan knura@yahoo.com wrote:
On Wednesday 02 Jul 2008, ravi kuamar wrote:
Then how do we find which hard disk had failed from the several hard disk in the server remotely. we need to hav certain command for this in linux.
Individual disks or hardware RAID?
If the disk is a member of a raid array, you can do a cat /proc/mdstat to know its health.
This will work only for sw raid. In the hardware raids I have come across the array is presented as a monolith device e.g. /dev/sda
If its an individual disk, I don't know of any direct, linux native way of finding it. Your hardware vendor might provide some daignostic software that can be run without rebooting the system.
If the disk does not show up in "fdisk -l " certainly rip'd. This is the easy one. If it is on it's way out, there will be timeouts or read errors.
But detecting such errors (OP) from a remote location. Hmmm, may be some kind of SNMP report?
-- Arun Khan
On Wednesday 02 Jul 2008 20:10, Arun Khan wrote:
On Wednesday 02 Jul 2008, Nadeem M. Khan wrote:
On Wed, Jul 2, 2008 at 3:29 PM, Arun Khan knura@yahoo.com wrote:
On Wednesday 02 Jul 2008, ravi kuamar wrote:
Then how do we find which hard disk had failed from the several hard disk in the server remotely. we need to hav certain command for this in linux.
Individual disks or hardware RAID?
If the disk is a member of a raid array, you can do a cat /proc/mdstat to know its health.
This will work only for sw raid. In the hardware raids I have come across the array is presented as a monolith device e.g. /dev/sda
If its an individual disk, I don't know of any direct, linux native way of finding it. Your hardware vendor might provide some daignostic software that can be run without rebooting the system.
If the disk does not show up in "fdisk -l " certainly rip'd. This is the easy one. If it is on it's way out, there will be timeouts or read errors.
But detecting such errors (OP) from a remote location. Hmmm, may be some kind of SNMP report?
remote syslog n nagios might be a lot simpler.
-- Arun Khan
ravi kuamar wrote:
In interview i was asked a question that. If a server is located at some remote place and in a server one of the hard disk had failed. Then how do we find which hard disk had failed from the several hard disk in the server remotely. we need to hav certain command for this in linux. in IBM AIX there is certain command but in linux i dont't know. does any 1 know this than plz reply.
For hardware RAID, you could use the RAID manager software. IBM's Raid Manager disks have RPMs for RHEL. You would have to hunt a bit to find the solution for other vendors.
-
Arun Khan -
jtd -
Nadeem M. Khan -
ravi kuamar -
Vivek J. Patankar