Hi, Lately I've seen a disturbing trend with qpopper - it freaks out and causes inetd to disable it:
Sep 3 11:42:41 brahe inetd[151]: pop-3/tcp server failing (looping), service terminated I have to kill -HUP inetd to get POP mail restarted. I only see these messages during the day, sometimes as frequently as ten minutes apart. The machine is a P133 128MB RAM, 3Com 3C905B, AHA2940UW with 2 Seagate Barracudas ... a pretty "normal" and stable setup. Kernel 2.0.35 with Solar Designer's security patches (the problem also occured with 2.0.32 and stock 2.0.35). The server currently has 1300 entries in the password file. I suppose it answers about 10-15 POP requests/minute, which doesn't seem out of hand. System load is about 0.15 on average, which seems acceptable :) Some other details: I grabbed the 2.52 source from qualcomm and saw the same problem - the deb was 2.3-4 from slink. For now, I've installed cucipop which seems to be better behaved, but doesn't do POP bulletins (which are very handy for ISPs). Anyone have any ideas for things I should look for? I don't see anything odd in the logs - there's not a common set of IP addresses or usernames retrieving mail around the time of a hang. IIRC there are some DSFG problems with qpopper 2.52, but I'd love to be of some help solving this problem, assuming others are seeing it. I don't think I'm qualified to write my own popper daemon, but then again, you have to start somewhere :) Kind of a rant, sorry. Private email welcome :) -- Nathan Norman MidcoNet 410 South Phillips Avenue Sioux Falls, SD mailto:[EMAIL PROTECTED] http://www.midco.net finger [EMAIL PROTECTED] for PGP Key: (0xA33B86E9)