Oh linuxadmins of Reddit, this junior linux admin has a confounding one.
I run a small cluster of servers for a club at my college that handle student projects and web-hosting for all other clubs on campus. One of my ubuntu boxen is in a weird spot. It has two nics living on the same subnet (campus sysadmin defined the IP structure via DHCP, even for static IPs):
Handles web-serving traffic, ssh gateway, and internal facing data movement to the VM cluster
10.0.130.9 10.0.130/24
Handles the software mirror
10.0.5.69 10.0.5/18
I have the route rules set to handle this sort of thing, but number 2 likes to forget it has an ip address about once every 2 weeks. There aren't any live cron jobs that would cause this and the monitoring system I use doesn't alert me fast enough to catch what the problem is in the kern logs before said logs are flooded with all the other messages generated by apache.
Any ideas on where I should be looking and how to approach this? I've tried what I can think of and google isn't being helpful.
TL;DR Ubuntu box is forgetful of second ip address and needs a swift kick.
Edit: formatting
[link][5 comments]