On 3/19/10, Jorge Armando Medina wrote: > Sasa wrote: > > If I have understood correctly the value of 'smtp-amavis' must be always > > equal or less of 'max_servers', in my configuration I have: > > > > $max_servers = 10; > > and > > smtp-amavis unix - - n - 5 smtp > > > > or pheraps these value must always equal ? > > Thanks. > > > > ------ > > > > Salvatore.
Since you are using Maia, they should be equal. If you were using amavisd-release (which you are not) a person may set $max_servers to one more than maxproc so there is a spare instance available for amavisd-release to use should all the others be busy. You are wasting memory if you load 10 instances of Maia into memory, and then only actually use 5. If you were at 2 before, I would not make such a large increase, as you may start swap thrashing (if you are not already). I would start by setting both to 3 (50% more than you had before) and then monitor the situation. It might also be a good idea to temporarily increase $log_level to 2 and then provide a sample or two of the TIMING line. filter:~# tail -f /var/log/mail.log | grep TIMING Mar 19 14:50:34 filter amavis[3633]: (03633-01) TIMING [total 478 ms] - sql-prepare: 13 (3%), SMTP EHLO: 10 (2%), SMTP pre-MAIL: 4 (1%), mkdir tempdir: 1 (0%), create email.txt: 1 (0%), sql-connect: 9 (2%), lookup_sql: 6 (1%), SMTP pre-DATA-flush: 5 (1%), SMTP DATA: 14 (3%), body_hash: 2 (0%), maia_connect: 6 (1%), maia_read_system_config: 3 (1%), maia_get_mysql_size_limit: 2 (0%), mkdir parts: 1 (0%), lookup_sql: 4 (1%), mime_decode: 30 (6%), get-file-type1: 135 (28%), decompose_part: 2 (0%), parts_decode: 0 (0%), update_cache: 5 (1%), maia_store_mail: 29 (6%), deal_with_mail_size: 4 (1%), maia_record_tests: 6 (1%), maia_delete_mail_recipient_reference: 4 (1%), fwd-connect: 47 (10%), fwd-mail-from: 3 (1%), fwd-rcpt-to: 5 (1%), write-header: 9 (2%), fwd-data: 1 (0%), fwd-data-end: 49 (10%), fwd-rundown: 19 (4%), main_log_entry: 37 (8%), update_snmp: 4 (1%), maia_delete_mail: 7 (2%), maia_cleanup: 0 (0%), maia_disconnect: 1 (0%), unlink-1-files: 3 (1%), rundown: 1 (0%) If you have not had problems before, disabling the cache will hurt rather that help. $enable_db = 1; $enable_global_cache = 1; What does the top of 'top' say? For example: top - 14:51:53 up 6 min, 1 user, load average: 0.14, 0.37, 0.23 Tasks: 63 total, 1 running, 62 sleeping, 0 stopped, 0 zombie Cpu(s): 0.3%us, 0.0%sy, 0.0%ni, 99.7%id, 0.0%wa, 0.0%hi, 0.0%si, 0.0%st Mem: 775728k total, 450604k used, 325124k free, 24340k buffers Swap: 240932k total, 0k used, 240932k free, 159924k cached -- Gary V ------------------------------------------------------------------------------ Download Intel® Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev _______________________________________________ AMaViS-user mailing list AMaViS-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/amavis-user AMaViS-FAQ:http://www.amavis.org/amavis-faq.php3 AMaViS-HowTos:http://www.amavis.org/howto/
