Re: I reinstall everytihng again ....

From: David Lord (snews_at_lordynet.demon.co.uk)
Date: 02/13/04 

Date: Fri, 13 Feb 2004 01:14:17 +0000 (GMT)

On Thursday, in article
     <1076617352.831595@ente.ipberlin.com>
     read_the_sig@do.not.spam.it "jpd" wrote:

> In article <20040212.1610.4500snz@lordynet.demon.co.uk>, David Lord wrote:
> > On Thursday, in article <1076596187.68810@ente.ipberlin.com>
> > read_the_sig@do.not.spam.it "jpd" wrote:
> >> In article <d312cd06.0402111626.1c8e1e3e@posting.google.com>,
> >> Reed Loefgren wrote:
> >
> > When I shutdown freebsd 4.7 I sometimes see something like the above
> > but don't have exact message and I'm not prepared to shut the server
> > down to log it.
> >
> > something like:
> >
> > syncing disks
> >
> > 99 99 99 99
> >
> > but only a few of them (possibly one per partition) then the message
> > that it has shutdown.
>
> 99 is not likely, the numbers are much more likely to be quite low,
> and often are diverse; AFAIUI they're buffers waiting to be flushed.
> Quite different from LILO faillure codes, and never seen on bootup.

It is definitely 99 repeated a few times at shutdown.

I just shutdown the Celeron 1300 box and had something like:

syncing (disks) 5

(crap memory - mine - I wish I'd written it down - but there was a
single 5)

> > ISTR when I had hardware problems the stream of 99 99... never ended
> > and plug had to be pulled. I replaced both hdd that certainly
> > appeared faulty and at same time took oportunity to replaced 24 MB
> > ram with 128 MB. The system (ISA+VLB+PCI m/b with AMD 586-133) is
> > still suspect as last time I tried I couldn't get a buildkernel or
> > buildworld to complete.
>
> Replace all the ram, clean the case and make sure its cooling is
> adequate (and it doesn't sit next to something hot). VLB slots are also
> suspect; take out the cards, thoroughly clean everything and re-seat
> them. Double-check the contacts and fix sternly. Then sacrifice a
> chihuaha under a full moon and try again. If it works, DON'T TOUCH.

In its previous 24 MB + becoming faulty hdd it had been up for months
then began rebooting every few days. The faulty hdd was 1.3 GB split
between swap, /tmp and /var so these were put on the 80 GB drive at
same time as ram swapped out. Everything was pulled and m/b, psu and
internals got a good cleaning but not the ultimate bath in hot soapy
water.

The m/b claims to support udma but I've had to set udma off in
loader.conf to get it to boot. It had been up for a month or more up
to New Year when it went down twice because of a couple of power
cuts. Since installworld at 4.7 It's been up 19 days so I'll try
cvsup to 4.9 and attempt the builds on this box. I'll try very hard
to note the shutdown messages in order to confirm the 99s.

David 

-- 
The Reply-To: is valid for at least 30 days after posting date
David Lord  - david@lordynet.demon.co.uk