How do I Recover from INed Accident in AIX 5.1

[JGutfreund]

The inevitable has happened. I was switching between modes in my application, which required me to drop to TCL mode in Unidata and su to root. At the unix prompt I got into some strange editor and typed a few things trying to exit and ended up killing the session, only to find that my server would no longer respond to telnets, ftps, etc.

Only later, I found, as luckily I still had another station logged in, that the /etc, /bin, and /dev folders had been renamed.

Scouring the internet for assistance I found this thread that explained it.

http://www.tek-tips.com/viewthread.cfm?qid=1081705&page=8

However, now I need to find a way to recover. I have booted into maintenance mode and mounted the hard drive, with the root folder, /dev/hd4 , at least that's what documentation indicates, but I can't do a list command to see the folders I need and I can't do a mv command to rename the folders.

Any assistance would be greatly appreciated. I've checked with IBM and since the O/S is so old, it's at end of life and not supported without a TTS ( Temporary Transitional Support ) SoW ( statement of work ), which probably equates to about $300 an hour in support.

Regards,
John

 

[sjm2]

cd /
echo *

What do you see? What have /etc /bin and /dev been renamed to?

 

[JGutfreund]

I'll have to boot into maintenance mode to see what is available, but the renames are as follows:

/dev is now /???dev
/etc is now /?etc
/bin is now /Qbin

 

[sjm2]

You said "I'll have to boot into maintenance mode to see what is available" - has the box booted as usual with the directories in this state? I am surprised. Anyway, you should be able to use full path names for /bin/ls and /bin/mv to fix these names.

 

[JGutfreund]

No, it was just hanging with LED code 518 initially and then went to 556. I found this document,

https://www-304.ibm.com/support/docview.wss?uid=isg3T1000132

and it indicated some corruption in the JFS. I followed the document and still no help. No matter what I did, I couldn't get root mounted enough to get to the /usr file system.

I ended up just doing a complete reinstall of the O/S to get the system running again. Back up and running now. I spent more time researching how to fix the issue than it took to rebuild the whole thing, so next time I'll just cut my losses earlier.

 

[DukeSSD]

Or keep a regular mksysb ready for recovery...