OK, I'm embarrassed to ask such a newbie question, but ... here goes. I run a small departmental POP server at work. It serves maybe, oh, around 70-100 people or so (just guessing). (So, to get ahead of myself a bit, we don't really have enough people - read: entries in "/var/mail" - for me to think that going to some two-level hash directory setup in the spool will help me with my problem very much. Read on ... )
It's running Qpopper 4.0.4 on a SPARCserver 20/71 with 128 Mbytes of RAM. (Yes, I know. It's vastly underpowered.) running Solaris 7 11/99. Under "normal" circumstances, the machine isn't really loaded. Relatively speaking, POP usage is somewhat light. But we have these few recalcitrant users who don't seem to know that "Keep Messages On The Server" is bad. Really Bad. Especially when these few people have mail spool files that are over 50 Mbytes in size. Some of them even closer to 100 Mbytes (or over). Naturally, every time these people POP in, the system goes into complete I/O and CPU starvation mode as all cycles get used up copying their huge mail spools to /var/mail/.luser.pop and then back again to /var/mail/luser. We have a 500 MHz Sun Blade 100 with 1 Gbyte of RAM running Solaris 8 2/02 in the on-deck circle. It'll run a full-fledged POP/IMAP/WebMail server with SPAM filtering, yadda yadda. But I haven't had time to get that running yet - too many other fires to put out first. (Always the story, isn't it?) In the interim, are there any steps I can take with Qpopper 4.0.4 to address this copy-to-.luser.pop-and-back-to-luser spool-file overhead problem? Thanks in advance. - Greg