11h Blog

Sometimes I wish that I could kick my own ass.

A long time ago, in a far away galaxy, I installed slackware on a hard drive. The thought was that I could use this drive as an emergency drive to power-on and back-up dying systems or resurect dead ones. So clever I was that I setup a very generic kernel to ensure that it would boot on any system. So clever I was that I changed the boot image from vmlinuz to vmlinuz.old and then changed lilo.conf to boot from vmlinuz.old. It took me five kernel compiles to catch on… Hint, the Linux 2.6 ‘make install’ copies bzImage to vmlinuz [but not vmlinuz.old]. What an idiot!

Offline. what?! My server, offline?!

“No way” I said to my self while traveling in France when trying to log into my server. There’s no way a fully redundant server has failed. UPS, dual CPU, hardware RAID5 with hot-spare. Down. Out. Not responding.

Thanks to Brian, I was finally able to move my file server out of the rack to a location where I could work on it. After a few minutes of debugging, I finally figured out why the server had failed:

Yes, I know that it is 2008. Yes, I know that ISA is extinct. I love old hardware. I like to recycle old things and bring them back to life. There’s no reason to throw something away simply because it’s old. Reuse, Recycle, reclaim!

For the last three years, this PCI and ISA CPU add-in board has been the backbone to my home network. It was a dual 550 MHz PIII with a 250 GB RAID 5 array with hot-spare.

It’s time to replace. One of the disks in the RAID failed in September and the controller immediately began rebuilding on the hot-spare. Now, it looks like the cap on the +5V rail on the ISA connector shorted, causing the D16 pad to burn itself right off the board. Glad to see that the PS was able to supply enough current to burn the pad off the board and NOT shut down…

Since this board is no longer functional, it is going to be reclaimed for its metal content.