Jul 14, 2003
Hey folks,

I have a relatively interesting problem that I'm having a heck of a time resolving.

On a server with over 100 virtual eth0: aliases, we're seeing several IPs dissapear off the map periodically, which naturally means the sites on said IPs go down.

Usually, just doing ifconfig (interface) (ipaddress) up will take care of the problem, but on several IPs, this does not help -- the IPs stay down regardless. If they show up in ifconfig's default output, then naturally downing them first then re-upping them works on some IPs, not others.

We've relocated some of the IPs to different machines, flushed the ARP cache on the router for it, and they work fine elsewhere; trying to ping using the interface, traceroute using the interface, or telnet using the interface dies at the machine (nothing to hop 1 on traceroute, pings dead).

The routing table appears to be updating properly on the local machine et al. Pinging on the machine itself to the specific IPs when ifconfig'd up will return properly, but otherwise, no access to the outside world.

Also, I've flushed iptables entirely, so it doesn't appear to be a firewalling issue.

What's funny about this is we've got a few machines pushing 200someodd virtual interfaces with no problems. OS is RedHat 7.3 w/latest patches et al.

Anyone seen this before?

Thanks :),

Oops, forgot:

WHM 8.5.1 cPanel 8.5.3-R2 RedHat 7.3 - WHM X v2.1.1
Last edited: