These forums are locked and archived, but all topics have been migrated to the new forum. You can search for this topic on the new forum: Search for solved-postfix error: cant create user output file on the new forum.
Web Hosting and Cloud Computing Control Panels
What does your /etc/procmailrc look like?
You may need to make sure that it contains these lines (amongst others):
ORGMAIL=$HOME/Maildir/
DEFAULT=$HOME/Maildir/
DROPPRIVS=yes
You don't have to delete the mail box file.
In my case, I could not do this because I would be deleting user's mail.
When I got this error, the issue was because someone had done:
[code:1]chown -R somesite /home/somesite[/code:1]
where "somesite" was the unix user corresponding to the virtualmin admin user for somesite.com. This of course changes the users of all the:
[code:1]/home/somesite/homes/someuser-emailuserA[/code:1] files and it messes up postfix - which expects to operate on a mail folder (/home/somesite/homes/someuser-emailuserA) with the same GUID corresponding to emailuserA.
It's not any more time consuming, I think, just to login to virtualmin for the virtual server having the issue, clicking the "Edit Email and FTP Users" and from the command line just issuing the following for each emailuser under /home/someuser/homes/emailuserA.
[code:1]chown -R emailuserA emailuserA/[/code:1]
The reason you login VM is to check the user names. Sometimes, when there are users with the same name on another VirtualMin instance, the system appends the website name to the user's name. like "johndoe.someuser.com" which changes the above to look something like:
[code:1]chown -R emailuserA.somesite emailuserA/[/code:1]
HTH!
its ok i worked it out, after checking and going though all the procmail configs...
it turned out some issue with the mail box/dir permissions...
for some reason virtualmin had only given them ownership to the master virtual server user rather than each of the individual email account users
ie helmet rather than admin.helmet
deleting the boxs and recreating them fixed it ...
also in other news if you tried activating bandwidth monitoring at some stage and maybe it didnt fully activate properly or you wanted to stop it but couldnt and see your syslog filling with this junk:
Jan 2 22:43:06 web kernel: BANDWIDTH_IN:IN=eth0 OUT= TOS=0x00 PREC=0x00 TTL=48 ID=58389 DF PROTO=TCP SPT=49415 DPT=80 WINDOW=65535 RES=0x00 ACK URGP=0
Jan 2 22:43:06 web kernel: BANDWIDTH_OUT:IN= OUT=eth0 *** DST=*** LEN=52 TOS=0x00 PREC=0x00 TTL=64 ID=49646 DF PROTO=TCP SPT=80 DPT=49416 WINDOW=65 RES=0x00 ACK FIN URGP=0
you need to delete /etc/shorewall/start and restart it
assuming you are using shorewall for your firewall
as it gets the following added to do the logging:
run_iptables -I INPUT -i eth0 -j LOG --log-prefix BANDWIDTH_IN: --log-level debug
run_iptables -I FORWARD -i eth0 -j LOG --log-prefix BANDWIDTH_IN: --log-level debug
run_iptables -I FORWARD -o eth0 -j LOG --log-prefix BANDWIDTH_OUT: --log-level debug
run_iptables -I OUTPUT -o eth0 -j LOG --log-prefix BANDWIDTH_OUT: --log-level debug
run_iptables -I INPUT -i eth0 -j LOG --log-prefix BANDWIDTH_IN: --log-level debug