At 15:15 2004-07-15 -0500, rgball wrote:
At my hosting service they use procmail to sort mail into all the various
files associated with the email aliases. I've been having lost mail problems
and have been looking at the procmail log to see if I can discover anything.
One problem I see on a very irregular basis is:
procmail: No space left to finish writing "rgb"
procmail: Truncated file to former size
BTW, the above local delivery failuer SHOULD result in a bounce to the
sender, unless your ISP is using procmail in a hokey fashion.
Other than the obvious (really running out of room) what might cause this?
It's the obvious one - running out of disk space.
These errors are interleaved with successful deliveries so I can't tell what
Keep in mind that most POP mail daemons (and even shell mail clients) make
copies of mailboxes when managing them (rather than running like a gap
editor), and thus someone accessing a very large mailbox can consume much
of the available space on a mount, because their disk use may be as much as
twice the size of their mailbox for a brief period of time.
Unfortunatley, shell clients generally don't run through the POP service,
and therefore when someone is accessing a huge mailbox via a shell account,
there may be absolutely nothing logged when they do, wherea a POP daemon
should log accesses to the maillog, so _remote_ users (or local users
actually using their shell client with POP) should have their basic access
A wise sysadmin should be able to focus on LARGE mailboxes and correlate
those to the freespace on that mount. I've seen people with mailboxes in
excess of a gigabyte, and that simply plays hell on mbox mail.
Sean B. Straw / Professional Software Engineering
Procmail disclaimer: <http://www.professional.org/procmail/disclaimer.html>
Please DO NOT carbon me on list replies. I'll get my copy from the list.
procmail mailing list Procmail homepage: http://www.procmail.org/