Hello,

This week, I was tasked with updating the Maintenance Level from 1 to 8 on a pSeries 650 server running Aix 5.2. I used the alt_disk_install method via smit. Upon reboot after the upgrade, the LED reported back 0518 /USR MOUNT FAILED. We couldn't even get console access. I tried a Serial Connection from a Laptop, a standard monitor to the graphics card - no dice. IBM actually suggest to reset the console by unplugging the mouse from the KVM and put it into the server directly / reboot / remove mouse cable and reboot again and we had the KVM console back up and I was able to get the system back online with 5200-01.

Has anyone had any similiar or other problems when going up in maintenance levels? Do I need to look at going to ML2 first then up?

Any help would be great.

Thank you.

I have exactly the same problem now!!

I'm moving from 5200-01 to 5200-07 ...

upon reboot, the system hangs with 0517 mount /usr...
I can't get into my AIX anymore...
I was even trying to do a preservation reinstall... but that won't work...
Right now I'm thinking to reinstall the entire system...

I don't know what happened with the update to 07, but it ain't good...

If anyone has any suggestions on this, i appreciate hearing them...

greetz

Are you on a p650 like me?
Did you do an alt_disk_install ? - If so you should be able to recover your system if you have access to the console and SMS - Set your boot device to the original device (in my case hdisk0)

no, right now i'm on this old machine (RS/6000 44P Series Model 170)...

I tried a lot of things whitin SMS. maybe indeed I should take a look at the bootlist.

Should I change it in the Limited Shell with the command "bootlist" or should I change it in SMS? ...

It will be monday when i'll be back at the office, so then i will take a look at it...

btw, is your system now in ML8??? or has it restored itself to previous state?

grtz

Hey guys,

I remember running into the 0518 LED code when doing some AIX upgrades on a contract this summer and looked through my email archive. I found this, hope it helps. I would guess maybe you have errors in /etc/filesystems that ML01 didn't care about but the new ML doesn't like . . .

Ralph


Wouldn't boot from 5.3 disk with error 0518 can't mount /tmp. When booting from 5.1 disk it didn't mount /tmp, just left you in the /tmp directory under root.

Found a line that read "nodename = oakbury" in the /tmp stanza of /etc/filesystems on both the 5.1 and 5.3 disk. After checking several other servers none had that. Did alt_disk_install -w to wake up altinst_rootvg then fixed it on the version of /etc/filesystems there. After that it would boot on the 5.3 disk.

Bottom line - if you ever get can't mount errors after a nim upgrade, check the /etc/filesystems file on your new disk for errors. Apparently some errors 5.1 let you get away with will prevent a 5.3 disk from being able to boot.

I had the same problem going from ML1 to ML7. Exactly the same issue which was resolved by going from 1 to 4 then 4 to 7; worked every time! This upgrade was on a couple of 44p machines.