On 4/10/99 3:04 AM, Colin Wood was rumored to have said...

>Greg Evans wrote:
>> So there I am..untarring something on my system and the next thing I know 
>> I see something about valloc() and I am in the debugger with a nice db> 
>> prompt...
>
>if this happens again, could you let us know what the output of typing
>'trace' at the prompt yields?

Yeah, sure thing..I just hope it never happens again...I should have 
thought to write that down.

as far as missing things...

traceroute
sysctl
syslogd
quotacheck
quotaon
update
portmap
are a few things I know are missing for sure

Sendmail  and named are now behaving funny with messages

named:
"<date> named[108]: cache zone "" (IN) loaded (serial 0)"
"<date> named[108]: master zone "sr-alpha.com" (IN) loaded (serial 
1999040100)"
 and a lot more like this 

then at the end
"<date> named[109]: Ready to answer quesries"

sendmail starts with named[109]: bad referral (1.1.168.192.in-addr.arpa 
!<*.168.192.in-addr.ARPA)

Neither of these happened before the crash a few mins ago...

> 
>> So I reboot, it goes through until it gets to filesystem check and has a 
>> big problem with my /usr partition, and tells me all about DUP BLKS and 
>> here I am now looking at removing things with messages like
>> 
>> DUP/BAD  I=75522 OWNER=bin MODE=100555
>> SIZE=1515 MTIME=Mar 30 20:23 1999
>> FILE=/sbin/ndc
>> 
>> So what the heck is this about and how bad did I just hose my system?
>
>it depends...what's not there anymore?  take a look in the lost+found
>directory for that partition and see if there's anything useful in there.

Only thing that is inthe lost+found for that partition is a directory 
called #101999 and a file called gettext.mo inside that, upon looking at 
that file it says 

"Convert binary .mo files to Uniform style .po files. Both little-endian 
and big-endian .mo files are handled......"


>> It 
>> _appears_ to be running fine so far, other than sysctl, and a bunch of 
>> other errors at startup...
>
>what kernel was this?  what kind of machine?

output from dmesg is 
NetBSD 1.3.3 (GENERICSBC) #0

The machine is a quadra 610, with 36 megs of RAM and a 3 gig drive, MacOS 
partition is the last partition

>
>> Arrrrrrrrrgh...where do I go from here? re-install?
>
>i usually do.

OK boys, how do I got about re-installing without losing all of the stuff 
I already ahve installed (e.g. my webserver and associated files which 
seem to have remained in tact) and of course the NAT sutff located in /etc

This also leads to my next question of upgrading...Once 1.4 is released 
is there a simple way to upgrade to 1.4 not including just re-installing 
from scratch?

Greg