E20 won't boot 1

[SeanAIX430]

Seems like I had a power failure this weekend. When I rebooted my 7024-E20 it did the memory check, scsi check but when it got to the boot check it gave me 3 long beeps and F91 showed up on the lcd and it went into the diagnostic screen that only let me test memory. I'm not having any luck finding an error message for F91, or the 00070000 that showed up in the one error log I could find. Think it's the system board or the scsi card? Thanks

 

[MoshiachNow]

F91 Low-level initialize VDISK file systems

You have a problem accessing your boot filesystems.
Typically, the problems with booting occur in the following areas:
1. Bootlv corruption
2. Filesystem corruption
3. JFSlog corruption
The following methodology is used to help repair/recover from the above named boot problems.
You will need bootable media (CD, tape) in order to insure the boot image is good. The level of the bootable media MUST be the same as the boot image on the installed machine.

1. Boot from AIX CD as if you do a Coldstart
2. Press F1 and enter to select the console (Key Board)
3. Type 1 and press enter for English
4. Type 3 and press enter for “Start Maintenance Mode…”
5. Type 1 and press enter for “Access a Root volume group”
6. Type 0 and press enter to continue
7. Type 1 and press enter to select hdisk0
8. Type 1 and press enter to Access this volume group and to start a shell. Now you should have a normal prompt.
9. At the shell prompt, we will try to do the repair mechanism. Enter the following:
fsck -y /dev/hd2
fsck -y /dev/hd4
fsck -y /dev/hd1
fsck -y /dev/hd3
fsck -y /dev/hd9var

Look at fsck output for any errors that are displayed. If a filesystem has been hosed, it may require a reload of the system, depending on the file system in question and the level of problem.
10. Assuming we have passed all the fsck's above, we will reformat the jfslog. If it has been damaged for any reason, it will cause the filesystems to not be mounted. Enter:
/usr/sbin/logform /dev/hd8
This command will ask the horrible question "Do you wish to destroy the log?". Just say yes. This will get you a new, clean jfslog.
11. Verify that the type field for the / (root) and /usr file systems is correct in the client's /etc/filesystems file. To do this, run the following command:

pg /etc/filesystems

Make sure that the / (root) and /usr stanzas contain a line that reads type=bootfs. If the type field is incorrect, use a text editor to correct it.
12. After this we will rebuild the boot image. Enter "exit" at the command prompt. This will complete the boot cycle by mounting all the file systems. Look for any error output, just in case.
13. At the prompt, enter:
lslv -m hd5
The output will show which disk has the boot image. Look under the PV1 column for this. Example output is below.

$ lslv -m hd5

hd5:N/A
LP PP1 PV1 PP2 PV2 PP3 PV3
01 0001 hdisk0
02
In the example, the boot disk is hdisk0. It is also vitally important that the data under LP and PP1 be both 0001! If it is not, then your boot image is not placed in the proper location on the disk and will need to be relocated!
14. Enter the following to recreate the boot image.
If /usr is not mounted – mount it as following:
mount /dev/hd2 /usr
Then run the bosboot:
/usr/bin/bosboot -a –d /dev/hdisk# , where # is the boot disk number.
15. Do a shutdown and reboot the system. This should clear up your problems. If not, detailed diagnostic information and support will be required.


Long live king Moshiach !

http://www.7for70.com/

 

[SeanAIX430]

Wow! thanks for all of the tips on fixing the boot files. It actually turned out that the boot sequence was wrong and when I pointed it back at the correct drive it worked again. The only other thing I did was replace the cable to all the drives, so it might have been that after all. Thanks a ton again!!