Let me tell you about my server. It’s a very basic Linux box with two hard drives mirrored via a RAID array. With the exception of the mail server, everything on the box is open-source, from mySQL to PHP, from the Apache Web server to the CMS that runs my business Web site.

My server is co-located at a company with a big, fat pipe to the Net. It has sat there for two years, churning quietly away without a hiccup.

Until about 6 p.m. yesterday. Of course, I didn’t know it hiccuped yesterday evening, since it was my wife’s birthday and we were out. It wasn’t until this morning when my phone started ringing that I found the server was down. I assumed at first it was a network problem, but a few hours of diagnostics proved that to be wrong. I couldn’t check the OS, since I couldn’t connect to the server. Somebody at the hosting company finally took a look. It said something about inconsistencies in files and suggested running FSCKK manually.

FSCK? It’s “file system check,” the Unix equivalent of Microsoft’s CHKDSK. It found four errors, which it fixed. Within five minutes, the server was up and running again—after about 20 hours down. Now I need to get my blood pressure down. Posting to this blog will resume as normal tomorrow.

I’m counting on another two years of seamless operation, but you know what my networking buddy says about the Net: It’s so complex it’s a miracle the damn thing works at all.