[rescue] Fixing the NVRAM battery in the SS5

[Bryan Gurney]

On Fri, 06 Jan 2006 16:12:53 -0500, Sheldon T. Hall  
<shel at tandem.artell.net> wrote:

> Quoth Bryan Gurney ...
>>
>> ... but still, this thing may boot up with a MAC address of
>> FF:FF:FF:FF:FF:FF or something, and refuse to continue.
>
> I haven't faced the problem, but I once saw a script that dealt with this
> once you got the thing to boot.  IIRC, you had to know the MAC address  
> and
> HostID, and put 'em in the script.  Obviously, you'd do that before the
> machine got so bad it lost its mind.  When the script ran, it stuffed the
> MAC address and HostID into the NVRAM somehow, set the clock, and  
> rebooted.
>
> I don't seem to actually _have_ the script, of course.


Well, either way, NetBSD installation is complete.  And this thing runs  
lean and mean after default installation; nothing like Debian!

load averages:  0.23,  0.13,  0.05                                      
13:18:46
14 processes:  13 sleeping, 1 on processor
CPU states:  0.0% user,  0.0% nice,  0.0% system,  0.0% interrupt,  100%  
idle
Memory: 25M Act, 116K Wired, 3784K Exec, 3280K File, 216M Free
Swap: 128M Total, 128M Free

   PID USERNAME PRI NICE   SIZE   RES STATE      TIME   WCPU    CPU COMMAND
   380 root      28    0   172K 1100K CPU        0:00  0.00%  0.00% top
     4 root      18    0     0K   11M syncer     0:00  0.00%  0.00%  
[ioflush]
   366 root      18    0   216K  960K pause      0:00  0.00%  0.00% csh
   346 root      10    0   188K 1808K wait       0:00  0.00%  0.00% login
     1 root      10    0    72K  848K wait       0:00  0.00%  0.00% init
   369 root      10    0   248K  820K nanoslee   0:00  0.00%  0.00% cron
   204 root      10    0   204K  636K mfsidl     0:00  0.00%  0.00%  
mount_mfs
   339 root       2    0  1184K 1704K select     0:00  0.00%  0.00% sendmail
   171 root       2    0   172K  852K kqread     0:00  0.00%  0.00% syslogd
   365 root       2    0    72K  808K kqread     0:00  0.00%  0.00% inetd
     2 root      -6    0     0K   11M sccomp     0:00  0.00%  0.00%  
[scsibus0]
     5 root     -18    0     0K   11M aiodoned   0:00  0.00%  0.00%  
[aiodoned]
     3 root     -18    0     0K   11M pgdaemon   0:00  0.00%  0.00%  
[pagedaemon]
     0 root     -18    0     0K   11M schedule   0:00  0.00%  0.00%  
[swapper]
# date
Mon Nov 21 13:19:35 EST 2005
# df -k
Filesystem  1K-blocks      Used     Avail Capacity  Mounted on
/dev/sd0a     7054670    344866   6357072     5%    /
/dev/sd0d     2064494       682   1960588     0%    /var
/dev/sd0e     8258270         2   7845356     0%    /home
mfs:204        507901         1    482505     0%    /tmp
kernfs              1         1         0   100%    /kern
#

14 processes, wow!  Running lean and mean.  As you can see, the time is  
still off (this thing isn't connected to the Internet yet; and I still  
have to go around somewhere and disable the LANCE ethernet since it has an  
SBus Happy Meal / SCSI card).  I probably went way overboard on  
partitioning the / partition, but we'll see.  At least too much logs and  
too much user stuff won't clog it up.  The partitioning isn't very  
disciplined, but I've seen worse (boxes with no separate /var or /tmp  
partition, and alerts coming in that / is full, and the cause being that  
there was "crap in /tmp").

There's no bash or tcsh, but at least ksh has tab completion for commands,  
as well as backspace support for the HyperTrm configuration I'm running  
now (ah, I remember the bad old days of trying to work fast on a sick  
Sparc box on a minicom crashcart terminal, botching command typography,  
and not being able to backspace).  And sshd isn't configured to start on  
boot, so I'll have to hunt around.  I'll have to RTFM, but first, I'll  
have to FTFM ("Find the filthy manual").