Server not starting after upgrading from 1.5 to 1.9

[vcp_ai]

I've tried to upgrade an old server running Proxmox 1.5 to last level with
# aptitude update
# aptitude safe-upgrade
Everything worked as expected , and I checked that grub menu.lst to point to 2.6.32 kernel. But after reboot:

1.- 2.6.32 kernel does not start, it says that Vmliuz.2.6.32... image is missing and grub fails with Error 15 File not found. Well, this is not actually a problem for me.


2.- When trying to use old kernel 2.6.18, it arrives to a point that enters some SXX-script (that I wrote in the past), it is waiting for some text files to 'remote start virtual machines' via rssh . This script perform a 1 sec sleep, looks for a file in a local directory, and if found, starts a vm machine via 'qm sart', and performs a echo "Loop number xxx" that showing in console. I am almost sure that this script is executed in background scriptXX & , but I'not 100% sure.....

Problem is that proxmox does not continue starting. Seems that network is not initialized, and I can not ssh to console, nor use web, and it does not arrive to offer a local login prompt.

Two vm machines that have autoststart, seem to start correctly.

Mi question is how can I obtain a login prompt to try to find/correct problem:
- There is no local virtual consoles (Alt-F1, Alt-F2, ..)
- Is there any virtual console supported via tty ??
- SSh is not started (Network is not started as I can not see any of the two IPs assigned to both NICs)
- I've tried to find if grub have some way, via edit command or similar, to force run level 1 to allow me to not start scripts in rc.d, without success.

Thanks in advance for your attention and solutions..

Vicente

 

[e100]

From grub you can use the edit mode and add the word single to the end of the kernel line.

kernel        /vmlinuz-2.6.32-6-pve root=/dev/mapper/pve-root ro single

If your script still runs when booting into single user mode then I would suggest using a live linux CD, mount the filesystem and change the files so your script does not run.

 

[vcp_ai]

Thanks for your answer.
Appending single to kernel params did not solved the problem.
Also tried init=/bin/bash without success.

So I mounted a live-cd GRML, and from a console terminal, isued commands picked on other post (about forgotten password)

short howto:
type 'lvs' to see your logical volumes (data, root, swap)
activate it with: 'lvchange pve -ay'
mount the root volume: mount /dev/pve/root /montaje
chroot into it: chroot /montaje
In /montaje we have our root filesystem
I renamed S99Manejador to K01Manejador to avoid execution.
exit, reboot, and all started to work

Thanks