Hi all,

I am running linux on IBM server
uname -a

Linux server1.example.com 2.4.21-27.EL #1 SMP Wed Dec 1 21:53:20 EST 2004 ppc64 ppc64 ppc64 GNU/Linux

In which one of the partition is mounted read only.
when i try to create folder in the corresponding partition i am getting this error.
mkdir: cannot create directory `test': Input/output error

/var/log/messages
EXT3-fs error (device sd(8,64)): read_inode_bitmap: Cannot read inode bitmap - block_group = 8192, inode_bitmap = 16777232

So i need to run e2fsck on that particular partition.But the partition size is 1TB and has more critical data.

so suggest me the proper e2fsck command to solve the error.And suggest me the proper method to do it.

I would recommend having a backup of this critical data somewhere else as a first step.
Do you know the dev name that is causing problems? You can type `mount` and find out what dev is being mounted for the partition you are referring to. For example: "/dev/sda2 on /home type ext3 (rw,acl)" means /home is being mounted from /dev/sda2.
I would then unmount that partition - example: `umount /home`
Then run `e2fsck /dev/sda2`

Hi zQUEz Thanks for you replay. just curious e2fsck /dev/sde is that command is sufficient or i have to send any parameters like -y . Bcos the e2fsck is going to run overnight. Just to know will formatting the partition will solve the problem?.

I unmounted the corresponding partition and run the e2fsck /dev/sde I got these errors



[root@server1 root]# e2fsck /dev/sde

e2fsck 1.32 (09-Nov-2002)

/dev/sde contains a file system with errors, check forced.

Pass 1: Checking inodes, blocks, and sizes Error reading block 268435458 (Attempt to read block from filesystem resulted in short read) while doing inode scan. Ignore error<y>? yes



Force rewrite<y>? yes



Error reading block 268435459 (Attempt to read block from filesystem resulted in short read) while doing inode scan. Ignore error<y>

do i have to continue with yes or no? And let me know do i want replace the HDD?

It is not sounding like your HDD is healthy at this point.
Assuming you have a backup, you could run with a -y flag to have it automatically assume yes to all prompts.
OR, you could use the -p flag which will have it fix errors that it finds that can be easily fixed, but if it comes across an error that may cause issues fixing, it will fail and stop.

I would say at this point, use the -y flag and see what happens.

I tried with -p flag but it stops with the error "Unexpected inconsistency". I have AMC for the HDD which is in SAN. so can i go for the replacement of running the e2fsck -y to solve the error. Bcos you have mentioned that HDD is not healthy.

Formatted the HDD problem solved.