Re: Updating OpenBSD




<jKILLSPAM.schipper@xxxxxxxxxx> skrev i en meddelelse
news:45853d15$0$90828$dbd41001@xxxxxxxxxxxxxxxxxx

As this box is located far away from where I am now, I am afraid that
rebooting now will make it not come up again. Am I right? Or can I simply
just reboot and run "rm -rf /usr/obj/*" and so again?

You can almost certainly just reboot. If you are really, really worried,
just unpack the version of baseXY.tgz you have installed again (using
tar xzpf, don't forget the p!).

Before reading your mail, I rebooted... and the box never came up. In the
car, drive 100 km... to find out that it was hanging during boot. It
actually went all the way through the dmesg, and ended with:

dkcsum: wd0 matches BIOS drive 0x80
wd1: no disk label
dkcsum: wd1 matches BIOS drive 0x81
root on wd0a
rootdev=0x0 rrootdev=0x300 rawdev=0x302

But then instead of starting to check disks, it asked for the location of sh

Well, I rebooted on the old kernel (which worked), did a "cp bsd.old bsd"
and then started from the top: cvsupdate and then building the kernel. No
errors - it went all through, untill I rebooted. Same error occoured. I
booted the old kernel again.

This will give you a new kernel and an old userland. Most of the time,
everything works, but I've seen problems - for instance, pfctl couldn't
load the rules because the kernel interface had changed. The box will
still come up, but be sure that SSH remains reachable!

I didn't even make it to the pf.

It is strange, because I have ("in serial") done this on two other OpenBSD
boxes where everything went fine. These two had identical hw, the failing,
third one has much newer hw.

Regards, Lars.


.



Relevant Pages

  • Re: The sad demise of an etch.
    ... investigating what is causing the badness in the current install. ... If I knew how to do this (besides a few variations on rebooting and ... kernel version and hardware). ... Rebooted to etch. ...
    (Debian-User)
  • Re: The sad demise of an etch.
    ... investigating what is causing the badness in the current install. ... If I knew how to do this (besides a few variations on rebooting and ... kernel version and hardware). ... Rebooted to etch. ...
    (Debian-User)
  • Re: Unpredictable performance
    ... as possible, rebooting between each run.. ... I'm going to run some tests without limiting the memory to 80 megabytes ... I can do some bisecting next week and see if I find any, ... The kernel is compiled as an UP ...
    (Linux-Kernel)
  • Re: The sad demise of an etch.
    ... investigating what is causing the badness in the current install. ... If I knew how to do this (besides a few variations on rebooting and ... kernel version and hardware). ... Rebooted to etch. ...
    (Debian-User)
  • Re: The crippled resurrection of said etch.
    ... investigating what is causing the badness in the current install. ... If I knew how to do this (besides a few variations on rebooting and ... kernel version and hardware). ... Saw it was starting gdm. ...
    (Debian-User)