Agent not seeing eth1

[cpowers]

The zabbix agent 1.1beta7 or 1.1beta8 do not see the eth1 interface on one of my servers. All my servers are running the same version of debian and there are only 2 noticable differences on this machine. 1. It is running the zabbix server and 2. It has a subinterface eth1:0, incidently there is no active eth0 on this machine because that NIC is not being used. The agent does get statistics for the loopback interface lo but neither network interface.

Has anyone seen issues with the eth cards starting at 1 instead of 0 or with subinterfaces on linux?

[crs9]

I run a lot of servers in Fedora Core 4. I recently started using zabbix and polling snmp info from all of them. I found a set of servers where basically eth0 and eth1 swapped positions in the SNMP list. It's only these two, the rest of the servers are fine. No idea how it happened or what. When I get a chance, I plan on swapping the hw address on the ethx devices to get them in the "proper" order.