[rescue] Trimming down Solaris 10
jon at jonworld.com
Wed Dec 3 19:55:13 CST 2008
On Wed, Dec 3, 2008 at 8:49 PM, velociraptor <velociraptor at gmail.com> wrote:
> It's a bit too spartan for things like an Oracle server (which
> requires X to install...*boggle*).
I know, it sucks!
> Right now I have a cobbled together franken-pkg-list in my profile
> resulting from me trying to fix the errors that resulted from $prev_SA
That's what I did at first, start with a 9 list and then figure out
what 10 needed until I got it to work on its own.
> Web consoles <shudder>.
I think part of what I have in there is a required dependency for zones, IIRC.
> Thank you for rescuing me from multiple iterations trying to fix my
> package list. /me hands Jonathan a cookie.
Yum, yum! :D
> For those of you with a lot of recent Sol 10 experience, what's the
> scoop on patching? smpatch, or some other patch-system-du-jour? Or
> just damn all their attempts to fix it and download the cluster and
> apply it the old-fashioned way?
It's been a year since I did any real admin work. We used smpatch with
some manual intervention. We had smpatch download what we needed and
then manually applied it with patchadd in single-user mode on reboot.
We had a S99patchme script in /etc/rcS.d (still works in S10) which
upon completion would rename itself so it wouldn't run again on the
next reboot. I *think* the patches were downloaded in the order they
were required to be applied for dependencies, so would do an ls -tr of
the /var/spool/patch directory and use that as the patch list for
patchadd. Again, my memory is fuzzy here.
More information about the rescue