View Single Post
  #2  
Old July 1st, 2008
john.samons's Avatar
john.samons Offline
Meditating on the code...
 
Join Date: September 2007
Location: Johannesburg, South Africa
Posts: 95
Re: Primary superblock is corrupt

Hi BVIJAY,

It's not quite clear. What is the problem you're experiencing?

AIX TL level and HACMP level please...

If you are staying that you can't do any file system operations? What do you mean...? Can you still write to the FS? Is production currently affected?

Or do you mean you cannot do operations such as dynamically increase the FS size...?

The FS isn't really under HACMP control, it's just managed by HACMP for failover.

It's also expected to get those fsck errors as your file system is currently mounted, which can cause ambigious output. For a proper fsck you have to unmount the file system first - and obviously before unmounting you would need to stop all processes accessing it, i.e. the application probabaly managed by HACMP.

Book some downtime on the system before you continue...

You could vi the /etc/filesystems file and set the check = true for this filesystem and failover to that system, however if you want a more direct approach....I would do the following...

---Assuming you have a failover cluster---
-bring the RG down
-this should stop your application, unmount the fs and varyoff the vg... if it's managed by HACMP
-varyonvg -c yourvg
then do fsck -y yourfs

Give us a bit more info and I've sure we'll be able to help more...





__________________
John Samons
Reply With Quote