This is not a general Postfix help forum - this is for the separate package PostfixAdmin which configures/maintains virtual domains & mailboxes using Postfix (plus MySQL or PostgreSQL, optionally maildrop, and Courier or Dovecot). For Postfix help you should head over to postfix.org and check out the Postfix mailing lists.
However, if all you want is to collate the LOGS into one place, then all you need is to configure syslog appropriately. Setup one server with inbound syslog connections permitted from the other servers, and configure each server to send logging to that one place. It's a bit tricky to do well with plain syslog, rsyslog is better, I prefer syslog-ng which had some advances filtering tools so you can do things like filter by source IP to (for example) put traffic from your bunch of mail servers into a separate set of mail logs.
Once you have the logs collated like that, it's simple to run whatever tools you want against them for reporting purposes.
However, collating the logging like that doesn't tell you the status of each postfix instance. For that you will need some form of monitoring (Nagios, SpiceWorks, etc). Ditto if you want to monitor things like queue length - you can either use something like Nagios, or for just visualising it you can use soemthing like Cacti (and some remote collection scripts) to gather the data, or use both. If you write (if there isn't one already) a Nagios plugin for queue length, then you can use the Nagios monitoring to also collect data (fed into RRD files) to visualise with Cacti.
As an aside, if you collect queue lengths into RRD files, you need to do a bit of work to combine them. Average is OK because average(a+b+c)=average(a)+average(b)+average(c) so consolidation works fine. But max(a+b+c is not the same as max(a)+max(b)+max(c) so you need to add the queue lengths together before doing any consolidation.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hey Folks,
I am searching for solution to collect multiple postfix servers log into a single centralized server to monitor overall postfix status.
Any suggestion?
Regards,
Singh
This is not a general Postfix help forum - this is for the separate package PostfixAdmin which configures/maintains virtual domains & mailboxes using Postfix (plus MySQL or PostgreSQL, optionally maildrop, and Courier or Dovecot). For Postfix help you should head over to postfix.org and check out the Postfix mailing lists.
However, if all you want is to collate the LOGS into one place, then all you need is to configure syslog appropriately. Setup one server with inbound syslog connections permitted from the other servers, and configure each server to send logging to that one place. It's a bit tricky to do well with plain syslog, rsyslog is better, I prefer syslog-ng which had some advances filtering tools so you can do things like filter by source IP to (for example) put traffic from your bunch of mail servers into a separate set of mail logs.
Once you have the logs collated like that, it's simple to run whatever tools you want against them for reporting purposes.
However, collating the logging like that doesn't tell you the status of each postfix instance. For that you will need some form of monitoring (Nagios, SpiceWorks, etc). Ditto if you want to monitor things like queue length - you can either use something like Nagios, or for just visualising it you can use soemthing like Cacti (and some remote collection scripts) to gather the data, or use both. If you write (if there isn't one already) a Nagios plugin for queue length, then you can use the Nagios monitoring to also collect data (fed into RRD files) to visualise with Cacti.
As an aside, if you collect queue lengths into RRD files, you need to do a bit of work to combine them. Average is OK because average(a+b+c)=average(a)+average(b)+average(c) so consolidation works fine. But max(a+b+c is not the same as max(a)+max(b)+max(c) so you need to add the queue lengths together before doing any consolidation.