recover from hard drive failure

[hong_3]

Lately I added an external drive to my box(11.0) and added it to the only
volume group of the system. Both /var and /opt file systems are configured
to use part of the new drive. Then the external drive started having problems. I had to take it offline to boot up the system. So now both /var and /opt are in a mess. Every now and then I see io error in writing. Is there anyway to clean up those two file systems, given that I have no backup tape?


Here is the syslog:
Apr 30 20:09:35 sis vmunix: vxfs: mesg 018: vx_idelxwri_done - /opt file system inode 28385 had a write error at offset 1015808
Apr 30 20:09:36 sis vmunix: vxfs: mesg 018: vx_idelxwri_done - /opt file system inode 28386 had a write error at offset 1015808
Apr 30 20:09:38 sis vmunix: vxfs: mesg 018: vx_idelxwri_done - /opt file system inode 28714 had a write error at offset 229376
Apr 30 20:09:39 sis vmunix: vxfs: mesg 033: vx_check_badblock - /dev/vg00/lvol8 file system had an I/O error, setting VX_FULLFSCK
Apr 30 20:09:41 sis vmunix: vxfs: mesg 018: vx_idelxwri_done - /opt file system inode 29060 had a write error at offset 163840
Apr 30 20:09:41 sis vmunix: vxfs: mesg 018: vx_idelxwri_done - /opt file system inode 29070 had a write error at offset 0

[Helen French]

Hi Hong:

It was not a good choice to remove the external disk which was already configured as part of the file systems. Incase if you want to remove any disks which is configure in the system, you need to remove the LVM information from the disk first. Otherwise the server and the file system will keep searching for the removed disk and will give you I/O errors.

Now, I would re-create the server either from ignite tape, or from backup or install fresh OS. This may save a lot of time. Otherwise, I would find out the problem with the external disk, rectify it or remove it from the system using LVM commands (lvreduce, vgreduce).

What is the error u getting when connecting the external disk ? If yoy can connect it back, then do it and find out the errors.

HTH,
Shiju

[pap]

It seems that after adding new disk and transfering /var/ and /opt to that disk you have not executed the following command before reboot.

#lvlnboot -R

you must have to execute it to update BDRA for root volume group. If you are failed to do so you get serious errors which makes the system unable to boot.

Now you can do the things right through
1. restore ignite image.
OR
Restore /var and /opt from backup.

-pap

[hong_3]

Shiju,

Here is the error in shutdownlog when I had problem with the external drive. Please advise. Thanks.


21:07 Thu Apr 11 2002. Reboot after panic: , isr.ior = 0'0.0'0
00:09 Fri Apr 26, 2002. Reboot:
22:14 Thu Apr 25 2002. Reboot after panic: SCSI: unrecoverred deferred error
22:21 Thu Apr 25, 2002. Reboot: (by sis!root)
22:42 Thu Apr 25 2002. Reboot after panic: , isr.ior = 0'14240001.0'9a268550
22:48 Thu Apr 25 2002. Reboot after panic: , isr.ior = 0'14240001.0'9a268550
23:32 Thu Apr 25 2002. Reboot after panic: , isr.ior = 0'14240001.0'735cd29c
23:56 Thu Apr 25, 2002. Halt: (by sis!root)

[Helen French]

Hi Hong:

Again, try these things:

1) Apply the latest patches to the system. Use the Custom Patch Manager from ITRC or install the latest patch bundle.

2) Check for HPMC errors. This can be checked from either the PIM information in boot console menu or from /etc/tombstones directory. If found anything wrong, then get it checked by HP support.

3) If you are able to attach the disk now, then try reducing the LVs. But, the best bet will be to restore from any type of backup.

HTH,
Shiju