Postfix setuid problem
 

Hi all

I run dist-upgrade from Debian Lenny to Squeeze. Three days monitoring & fixing all *million* little errors (mostly php and apache related) and seems that everything works again like a Buick!

Only errors that I found now is from mail.log

Error 1

May 9 09:21:55 mail postfix/proxymap[26633]: fatal: setuid(1003): Resource temporarily unavailable <--- UID 1003 is postfix
May 9 09:21:56 mail postfix/master[26595]: warning: process /usr/lib/postfix/proxymap pid 26633 exit status 1
May 9 09:21:56 mail postfix/master[26595]: warning: /usr/lib/postfix/proxymap: bad command startup -- throttling

Error 2

May 9 11:10:06 mail postfix/smtp[28594]: 47D3B1814E9: to=<www-data@mail.domain.tld>, relay=none, delay=0.01, delays=0/0/0/0, dsn=5.4.6, status=bounced (mail for mail.domain.tld loops back to myself)

Error 3

May 9 10:38:18 mail amavis[26828]: (26828-12) (!)FWD via SMTP: <newsletter@some.company.tld> -> <some@person.tld>, 451 4.5.0 From MTA([127.0.0.1]:10025) during fwd-connect (Negative greeting: at (eval 101) line 555.): id=26828-12
May 9 10:38:18 mail postfix/smtp[28280]: 0197F1814E9: to=<some@person.tld>, relay=127.0.0.1[127.0.0.1]:10024, delay=37, delays=0.02/0.01/0/37, dsn=4.5.0, status=deferred (host 127.0.0.1[127.0.0.1] said: 451 4.5.0 From MTA([127.0.0.1]:10025) during fwd-connect (Negative greeting: at (eval 101) line 555.): id=26828-12 (in reply to end of DATA command))


Many installation guides on the Internet says that you have to remove libio-socket-inet6-perl but does not describe why? (I got that library installed on my system).
So far any hit from google does not have solved these error messages, however mail server works; I can receive and send mails via It just fine.

I have not made any changes to postfix, spamassassin, amavis or clam-av configurations on this machine, all have been worked without those error messages on Lenny and works now but with errors on mail.log

I'm guessing that I need to add/tune some configuration "things" or completely change things but can't find anything that could cause the errors so it's hard to try or change anything..

Could someone give me a hint or two? :)

not sure for the hint but libio-socket-inet6-perl is for IPv6 and here in logs you are using only IPv4. so probablly it's not effecting something.

Strange.. my /etc/security/limits.conf was empty before (as it should be) but now I added these lines

postfix soft nproc 45
postfix hard nproc 60


Error 1 and Error 3 disappeared.. This machine is virtual machine and I can't compile my own kernel so maybe the providers new kernel got something "new and super" on it?

found also this from dmesg

[ 0.676125] SELinux: Initializing.
[ 0.676143] SELinux: Starting in permissive mode

but that should not be the problem as it is in permissive mode

Now just need to wait and see did the errors really go away. I did not remove libio-socket-inet6-perl library.

Seems that the problem was that nproc thing (I increased that amount more), and I also put some aliases that routes @mail.domain.tld mails to postmaster @domain.tld. There is no more errors left in the mail.log