What to do?

[Paul Civati]

"Ian Hunter" <ihunter at hotmail.com> wrote:

> I'm having lots of bother with v3.1.
> 
> Basically, since I installed it, it has had infrequent panics and reboots, 
> 
> 1. I've installed the WC 3.1 from CDROM. Is there any reason why I should 
> upgrade, and if so what to?

3.1 is early in the 3.x branch and my understanding is that those
early releases were rather 'rough', I use 3.4 here on my workstation
with softupdates and it's rock solid.

Perhaps check the 3.1 errata and the 3.2, 3.3 and 3.4 release notes
to see if any mention was made of fixing any panics.

> 2. My machine has 32M - is this adequate? (In my day..... ;-)

Workstation or server?

I ran a 32M workstation using 2.2.x for about a year, usable for about
upto 10 xterms but slow.

I also ran an OpenBSD sun4m machine for 2 years with 32M, again usable
but sometimes slow.

But hey, memory is cheap now. :)

My new xciv.org server has 128M and after 100 days uptime:

Memory: Real: 9572K/72M act/tot  Free: 44M  Swap: 4K/512M used/tot

My home workstation only has 64M because the 440TX chipset won't
cache >64MB otherwise it would be more by now.

My work workstation has 128M, the only problem of course is netscape
likes to grow:

Mem: 88M Active, 8788K Inact, 18M Wired, 4488K Cache, 8187K Buf, 4508K Free
Swap: 512M Total, 66M Used, 446M Free, 13% Inuse

> 3. The majority of these crashes leave me unable to determine what has 
> happened - it reboots and smiles at me :-). Is there somewhere I can look 
> after the event?

If configured and you have enough space you should get a crash dump
in /var/crash, you can then I think use dmesg with some options to
get the dmesg against that crash dump file(s).

Usually the panic should also get logged, depending on whether you've
changed your syslog configuraion.

-Paul-