We should collect some basic postfix metrics to help us spot trouble, at least queue size and message delivery/failure rates.
Description
Description
| Status | Subtype | Assigned | Task | ||
|---|---|---|---|---|---|
| Resolved | Jgreen | T91508 [Epic] overhaul fundraising cluster monitoring | |||
| Resolved | • cwdent | T152562 Port fundraising stats off Ganglia | |||
| Resolved | • cwdent | T176495 prometheus collector or exporter for postfix metrics |
Event Timeline
Comment Actions
For posterity, we now have delivery rates added by log-scraping:
class { 'prometheus::collector::syslog':
jobs => {
'/var/log/mail.log' => {
'mail_bounced' => 'postfix\/smtp.*, status=bounced',
'mail_deferred' => 'postfix\/smtp.*, status=deferred',
'mail_expired' => 'postfix\/qmgr.*, status=expired',
'mail_sent' => 'postfix\/smtp.*, status=sent',
},
},
}Comment Actions
Timing-wise, it looks like this broke on Jul 12, 2017 between 15:59:01 and 16:02:29 UTC.