Server downtime...

Started by NFG, January 09, 2002, 01:02:09 AM

Previous topic - Next topic

NFG

This was written before I fixed it, but you couldn't read it.  Solution's at the bottom, and no I didn't have to fly to Japan.  Can't decide if that's a good thing or not...

--

So my server went completely down. Several personal domains, documentation for our customers, email and FTP are all unavailable because of some mysterious network failure. Normally this wouldn't bother me, but the server's in Japan and I'm in Australia, and the only way to fix this problem is to go there and take care of it.

Of the four computers on the LAN, only one can connect to the internet. It's not a problem with the switch, as we've swapped ports to no avail. It doesn't appear to be a problem with the router, as I can connect to it and verify that every setting is correct, and it does still let one machine onto the net (unfortunately it's not a server and has no serving capability).

When I remotely connect to the last available machine, it cannot ping or otherwise access the other three machines on the LAN. For all intents and purposes these three machines have simply dropped offline simultaneously.

We've bypassed the switch, swapped cables and ports, and rebooted everything a million times. I don't suspect the machines, because it's basically unheard of for three machines to die at the same time (one of which, a laptop, wasn't even connected or plugged in at the time the other two died).

I don't suspect the switch, 'cause we've bypassed it entirely to no avail.

It's not the modem, 'cause the router works fine.

That leaves the router. I have trouble suspecting the router because one machine works just fine kthx, and why would it suddenly deny access to every other machine but not that one? It's a frickin' mystery.

Before flying back to Japan next week to fix it, I'm going to have the router replaced and hopefully reconfigure it and see if that helps. If it does, I'm going to set the old router on fire and scream for an hour.

--

Turns out the problem was a series of coincidences all in a row.  Neither server would boot, and while I still don't know why one of them was down, the other was hanging on the BIOS screen begging someone to please, please press F1 just for a moment so it could load up.

We ended up buying a new router, which didn't help 'cause the logon info we had for the fibre was outdated and was logging into a dynamic IP instead of the fixed one, so while our tech support in Japan (Zumi's mom) reported it was working, we couldn't access the network.

So we've got one dead server, one spare router, and I'm buying and building a new computer next week to be the primary server, with the current one as backup.

<sigh>

A week, a whole week (well, six solid days) without my servers is way too long.  I actually had to go outside.