How to check the harddisk bad sector

[Chui Kock You_1]

Hi,

What is the command line to check the hdd whether in good condition or not. Got any bad sector or not. The server is L2000 running on hpux 11i version 1

Regards
Chui

[Ivan Krastev]

Try to read data from the disk with dd.

regards,
ivan

[Torsten. · Accepted solution]

This will help you:

http://docs.hp.com/en/5991-1236/When_Good_Disks_Go_Bad_WP.pdf

[SKR_1]

Use below command

dd if=/dev/rdsk/cxtxdx of=/dev/null bs=2048

Thanks

SKR

[Asif Sharif]

Hi Chui,
see the below thread.

http://forums11.itrc.hp.com/service/forums/questionanswer.do?threadId=1061965

Regards,
Asif Sharif

[Steven E. Protter]

Shalom,

The bad disk doc is top notch.

I'd do this:
dd if=/dev/dsk/blahblah of=/dev/null count=10000

Raise the count base don disk size.

This will find a bad sector.

SEP

[Sandeep_Chaudhary]

u can use dd command

Exit values are:

0 Successful completion.
>0 Error condition occurred.

DIAGNOSTICS
Upon completion, dd reports the number of input and output records:

f+p records in Number of full and partial blocks read.
f+p records out Number of full and partial blocks
written.

When conv=block is specified and there is at least one truncated
block, the number of truncated records is also reported:

n truncated records

EXAMPLES
Read an EBCDIC tape blocked ten 80-byte EBCDIC card images per block
into an ASCII file named x:

dd if=/dev/rdsk/c0t0d0 of=/dev/null count=1000

unix # dd if=/dev/rdsk/c3t15d0 of=/dev/null count=1024
1024+0 records in
1024+0 records outhis disk does not have bad sector as records in and out are matching


means t

[Steven Schweda]

My favorite method is to attach the disk to
a Solaris system, and use the "format"
program there. It can do some fairly
low-level work with defect lists, and it can
easily do some multi-pass, multi-pattern
write+read testing.

[Adam W.]

I use ioscan -fCdisk, to ensure disk is claimed and identified correctly, so then I run, diskinfo /dev/rdsk/cxtxdx, again to see if the disk is identified correctly, so for further investigation I normally run dd if=/dev/rdsk/c0t4d0 of=/dev/null bs=1024k count=64, and a dd if=/dev/rdsk/c0t4d0 of=/dev/null bs=1024k count=128. Now, the 64 and the 128 SHOULD both came back with 64 records in/out and 128 records in/out, if you have a good disk.

[Chui Kock You_1]

Hi,

If i used the command dd if=/dev/rdsk/c0t4d0 of=/dev/null bs=1024k count=128 will it corrupt my existing data or not.

Regards
Chui

[Vivek Bhatia]

Hi,

Use the following manual for checking the disk integrity of the disk.

www.docs.hp.com/en/5991-1236/When_Good_Disks_Go_Bad.pdf

This is a Very helpful document.DO have a look

Regards
Vivek Bhatia

[likid0]

Nope, your data won't be harmed, beacuse you are only reading from the disk(if=) and writing to /dev/null (of=/dev/null).

You can execute it without problem.

[Steven Schweda]

> You can execute it without problem.

One of the nice features in the Solaris
"format" program is its ability to do surface
analysis using this sequence:
save - write - read+verify - restore.
It lets you do write+read testing without
losing the data (unless the disk is bad). It
also has data-destroying options:

format> a

ANALYZE MENU:
read - read only test (doesn't harm SunOS)
refresh - read then write (doesn't harm data)
test - pattern testing (doesn't harm data)
write - write then read (corrupts data)
compare - write, read, compare (corrupts data)
purge - write, read, write (corrupts data)
[...]