Greetings

Hope somebody can help here please and to be honest I am embarrassed to ask such a simple question =/ - However the RedHat Reference Material only seem to supply information about surface scans during the install phase. :/

Running RedHat 9 - Kern 2.4 - up and running (post install)

Target drive/partition (ext3 - Primary partition 10 GB) is /dev/hdc1 (IDE Secondary Master)

All I would like to do is run a disk check (surface scan) on this partition ( hdc1 ) - I'm sure that this is a one liner shell command that has evaded me.

I would prefer a destructive write test over a simple surface read test but anything will do!

Cheers for any input!

Svha

Assuming that is not your / partition, you'll want to umount it first:
umount /dev/hdc1
Then run:
fsck -f /dev/hdc1
The -f forces it since it's likely marked clean.

If you want to make it easier:
shutdown -Fr now
Will force a fsck upon reboot.

Cool

Perfect! - Thanks MasterC =)

What I have found the most difficult switching to Linux are some of the command line options - in this case;

fsck = file system check perhaps?

Logical once you know! Therefore the list of "Commands and their explantions" listed in your footer will be an absolute gem!

Many Thanks ..... Svha

If you are wanting to do a "surface scan", do you mean that you are trying to find badblocks, rather than just testing the file system?

If so then you need to man e2fsck and look at -c

-c This option causes e2fsck to run the badblocks(8) program to
find any blocks which are bad on the filesystem, and then marks
them as bad by adding them to the bad block inode. If this
option is specified twice, then the bad block scan will be done
using a non-destructive read-write test.

Corin you just hit the nail on the head!

I was a little untrustworth of the drive at first but it seems happy (so far!)

Yes I wanted a complete surface scan but from what I have learnt about IDE drives recently is that they will only keep the bad block register within the operating system, once the operating system / filesystem is destroyed you may aswell move back to square one!

Which is perhaps why I have ran 4*9GB SCSI drives (chucked in the bin at work) on my main box for quite a while (nearly 3 years to be exact, maybe Mylex controllers just dont like any hard disks what so ever! it seems! (no further comment!) hehe ). My Adaptec 2940UwPro SCSI controller has spent many an hour finding "Drive Defects" - upon these drives and remaped them through the Adaptec BIOS/Firmware - And afterwards they perform perfectly! (Call me out dated but a i686 P700 and 28GB of storage is enough for me)

However it seems with IDE drives
That the bad blocks can only be remapped by software and or "an intelligent operation system"

Is this still true with newer IDE drives? I remember Compaq selling a large range of "SMART" IDE drives / controllers / BIOS - that did actually did detect errors on the IDE channel - but still refused to boot with a defective IDE drive.

Anyhows thread closed!
My next door neighbour is impressed with his new 10GB drive (thats 5 times bigger than his old 2Gb drive) and it seems to be working ok - or as well as a Windows proxy server could work! ( /me hopes NTL will not notice the RJ45 cable running from his bathroom window to the neighbours proxy server)

Cheers and thanks for the input folks!

Svha!