Older blog entries for neuro (starting at number 3)

Woke up at 5AM to hear the alarm system beeping. A quick look at the clock confirmed that the power was indeed off. I went back to sleep and let it beep.

Woke up at noon and went to see how things made it. The power was off for about an hour, so the UPS didn't last through the whole thing. :( My Slackware 3.1 inplace-upgraded-to potato system had a few things in the inittab pointing at the wrong place, so it had refused to boot. fsck on a 20GB drive on a 2.0 kernel takes a long, long time. Cleaned it up, and got it to boot, but no networking was setup, so I had to setup an /e/n/i for 2.0, and reboot again.

After all that, I finally had it up and running, other than snmp. mrtg started sending me mail every 5 minutes, until I setup the potato snmp, rather than the libc5 one I was running before the reboot. The reboot seems to have fixed the very weird dates I was seeing in ps, however, so maybe it was a good thing.

Tried to go get a haircut, but was too late, already closed. Came back and worked on networkmod a bit.

Tomorrow seems like a very good day for nethack.

/e/n/i support isn't into potato or woody yet, but is written. Apparently it is still not perfect with saving comments either, tho.

www died several times in the morning. Someone had forgotten to turn on ECC in the BIOS. After enabling it, 2.2.15 still had problems. Went back to 2.2.14, and got an NMI at boot. Ripped RAM out of machine, and found someone with some spare, and put that in while getting the dead one RMA'd...

Still didn't get the installer source code online, but I did get an rsync-push of the internal cvs server set up to be available outside the internal network, and got anonymous cvs and cvsweb setup for it at cvs.stormix.com.

Built updated boot disks again...

And now the weekend starts!

I've noticed some corruption in my "notes" section, and others in the projects section, so people may want to watch for other weirdness like that...

Looked at linuxconf in woody and potato and didn't see anything dealing with /e/n/i, so maybe I wasn't reinventing the wheel. Send some mail off to the maintainer to find out more...

Built updated boot disks, fought with them to boot on the SCSI Athlon, and when successful, found out another option was needed in the kernel, which is already tight on space. Fun fun fun.

Didn't get the installer source put online (with the fun on www) as I was intending, so it is up for tomorrow

It seems 2.2.15pre17 is just not liking the hardware of www. Processes segfault randomly, the network card goes off into one second 300ms away, the next .1, on a one hop all 100Mb FDX connection. Something is weird. I am beginning to suspect the RAM, which may not be ECC. It was running fine for awhile after a reboot, but it didn't last long enough for me to drive home :(.

got a reply from mkp, and he's busy busy on xfs, but it seems the gnome BTS doesn't send him e-mail on bugs, so he's never seen the patches for GDM. Subscribed to the gdm list, and will be resending them if the selected portions haven't been heavily rewritten.

Finished off the work box of penguin mints. Only one spare left, so out goes another order for more...

And yah, this diary entry is in the wrong day, because it was a lonnnng day.

Finally decided to see what everyone was talking about. It's interesting that I was thinking of setting up a diary myself, only to find that there is already one!

Today the work on the SAS Network module continued. For those that don't know, SAS is the Storm Administration System, which is getting a lot of big changes for the next release of Storm Linux 2000. Talking to aj it seems that linuxconf may have something to look at as well for the backend interface to /e/n/i

Not much else interesting in this first entry...

PS: Culus looks quite evil with red eyes...