[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Strange problem




> On Fri, 2002-02-01 at 00:59, Jeremy C. Reed wrote:
>
> > In my experience, unstable is "unstable".
>
> At times.

We have production boxes running unstable with no problem. Needed to run
unstable because only unstable had some new software, unavailable in
stable. Its a pity stable gets so outdated all the time as compared to
other distros like Redhat and Caldera (stable still on 2.2 kernel), but
thats a topic for a separate discussion.

>
> > > and came up in a very strange state. No users could log in, only
root, and
> > > things like ps, w, and top wouldn't work. I was called, got in via
ssh,
> >
> > Why happens when you runs these commands? (What does "wouldn't
> > work" mean?)
>
> They hung. Nothing happened until I hit ^C
>
> > What do the logs say?
>
> Nothing. syslogd is one of the things that didn't start.
>
> > What do you mean that it is impossible to be the same? (Are you saying
> > that proc was also mounted at / ?)
>
> Hmm. I didn't say that right. Mount showed /proc mounted. 'df' *also*
> showed /proc mounted, with the same size/used/free as /.
>
> >
> > > /proc by hand, started up the utils that didn't start, checked
things out
> > > the best I could, and rebooted again. Same thing. I've gone through
> >
> > What do the kernel messages say?
> Nothing.
>
> > What do the logs say?
> Nothing

Well if syslog isn't started... no wonder. Is it possible to hand-start
syslog after the box has started up, to have it record further error
messages once it the box has booted up?

> > What are these utils that didn't start? (Some network services that
need
> > to be correctly setup in /etc/rc*.d/ ?)
>
> networking, syslog, just about anything that needs /proc to me mounted
> and readable.
>
>
> > Sometimes when I upgrade from stable to unstable, I have had some
packages
> > not reinstalled and some software didn't start that should have.
>
> This machine was running unstable for quite some time, stabley. :)
>

Yeap, for us too. Although we always upgrade a non-critical box first to
see if it breaks anything.



Reply to: