A couple of days ago, xinetd on one of two RH 8.0 machines began running at 96%+ CPU. The other is at about 0.6%. This started while I was screwing around with things while these two and two Win2K machines had two NICs each. I have since removed all of the extra NICs but cannot get xinetd to behave.

The network is up and everything else seems to be working (well swat isn't but samba is). I am clueless. Can someone suggest what to look at to get xinetd to settle down?

It could be a lot of things. Exactly what were you "screwing around" with? I treat these situations like I lost my keys. Backtrack to where you remember things working fine.

Unfortunately, things have not "worked fine" on this particular machine for several weeks. I should have said that I noticed the xinetd problem a couple of days ago - don't really know when it started for sure. The last time things were really fine, I was running RH7.3.

A couple of days ago, I used BIOS to disable the NIC that had been eth0. RH reactivated the disabled NIC somehow. This caused a series of problems that resulting in my using the RH network configuration tool to delete and then add back first one and then a second NIC. Then I got messages about conflicting aliases for the NICs.

Today, I physically removed the card and RH decided that I had no NIC at all (it was activated in BIOS at this point). Xinetd still freaked. I could no longer find files related to eth0 configuration on the system but have another RH installation on a machine with the same NIC driver. As root, I created an identical ifcfg-eth0 in /etc/sysconfig/devices, restarted networking, and could then connect to the LAN and Internet again. Xinetd is still monoplizing a CPU and now I cannot edit eth0 with the RH Network Device Control Tool. The permissions on ifcfg-eth0 are the same as on the "good" machine and I can edit that one with the tool.

What is xinetd looking for on this machine? I know it has some debugging and logging options but am too much of a newbie to know how to turn them on.