Because of user access growth, the number of processes associated with IMAP has increased and I thought to change some of the configured parms. dovecot -n yields:

# 1.1.14: /usr/local/etc/dovecot.conf
# OS: AIX 1 005A928C4C00 listen: *:143
ssl_listen: *:993
disable_plaintext_auth: no
verbose_ssl: yes
login_dir: /var/run/dovecot/login
login_executable: /usr/local/libexec/dovecot/imap-login
login_processes_count: 12
login_max_processes_count: 774
max_mail_processes: 1280
mail_max_userip_connections: 12
verbose_proctitle: yes
first_valid_uid: 200
mail_location: mbox:~/mail:INBOX=/var/spool/mail/%u:INDEX=/var/dcindx/%u
mbox_write_locks: fcntl
mbox_dirty_syncs: no
auth default:
 passdb:
   driver: pam
 userdb:
   driver: passwd

The only changes was that max_mail_processes went from 1024 to 1280. Now I get a error message when I start DC:

Warning: fd limit 2000 is lower than what Dovecot can use under full load (more than 2054). Either grow the limit or change login_max_processes_count and max_mail_processes settings

So I changed the no_size and no_size_hard to 3500 and 4000 respectively in both 
dovecot and root

AIX defines nosize:* *Sets the soft limit for the number of file descriptors a 
user process may have open at one time.

an lsuser dovecot returns:

dovecot id=417 pgrp=dovecot groups=dovecot shell=/bin/false daemon=true admin=false ... fsize=2097151 cpu=-1 data=262144 stack=65536 core=2097151 rss=65536 nofiles=3500 nofiles_hard=4000
I kill dovecot and all children and restart it....same error message

What am I missing? Does the machine have to be rebooted for the no_size to be updated?


--
==== Once upon a time, the Internet was a friendly, neighbors-helping-neighbors small town, and no one locked their doors. Now it's like an apartment in Bed-Stuy: you need three heavy duty pick-proof locks, one of those braces that goes from the lock to the floor, and bars on the windows.... ==== Stewart Dean, Unix System Admin, Bard College, New York 12504 sd...@bard.edu voice: 845-758-7475, fax: 845-758-7035

Reply via email to