Hello, I know this has been discussed before, but I still have not found a way to remedy this problem. Users on slow connections with large/lots of messages end up having problems retrieving their messages. The following entry is logged. (xxxx in place of host/username, etc. for privacy)
-------- May 26 09:43:58 xxxx popper[9100]: I/O error flushing output to client xxxx at xxxx [xxx.xxx.xxx.xxx]: Operation not permitted (1) [pop_send.c:689] -------- 90%+ of the users are using Outlook Express...most of the remainder are using Outlook...some Eudora, Netscape Mail, or Mac versions. My setup: RedHat Linux 9.0 Dual Intel(R) Xeon(TM) CPU 2.40GHz (hyperthreading...Linux sees as 4 CPUs) 2GB RAM, 2GB swap. Attached devices: Host: scsi0 Channel: 00 Id: 00 Lun: 00 Vendor: Promise Model: 2+0 Stripe/RAID0 Rev: 1.10 Type: Direct-Access ANSI SCSI revision: 02 Qpopper version 4.0.5 (standalone) Current configure options: ./configure --enable-poppassd --enable-shy --enable-chunky-writes=1 --enable-hash-spool=2 --enable-temp-drop-dir=/var/spool/maildrop --enab le-standalone --enable-debugging /var/spool/maildrop is in a different partition than /var/spool/mail Started at boot (rc.local) with /usr/local/sbin/popper 110 -scC -T 600 The server does not appear to have performance problems because of popper, but users that get these errors get frustrated. Sometimes they can wait a few minutes and retry and it works...sometimes we have to remove large messages from their mailbox. I have tried a couple of other POP3 daemons (cucipop, popa3d), but I haven't found one that works well, uses hash mail dirs, doesn't give the users duplicate messages if they use pine as well as Outlook*, etc. I would gladly provide additional information about my setup and this problem if it would help get me pointed in the direction of solving this problem. Thanks! Curt
