Monitoring guidelines for SurgeMail
The following are
guidelines for monitoring a SurgeMail system. Surgemail has
built in monitoring, restart, and reporting for most critical
issues.
Use the following
settings to automate monitoring...
- g_manager "your@email.address" - Define the address to get
important reports.
- g_orbs_report
"your.ip.address" - The ip address of your server, used to
check if your server has become blacklisted.
- g_pop_warning "400"
- Sends report if more than this many POP or IMAP sessions
occur. Set at appropriate level for your system.
- g_smtp_warning "200"
- Send report if this many smtp sessions at one time.
- g_queue_warning
"10000" - Send report if more than this many outgoing messages
are queued.
- g_user_send_warning
"5000" - Send report if a single users sends more than this
many messages in a day (likely spammer)
- g_lowdisk_warning
"5000mb" - Send alert if free disk space drops below this
level.
- g_inbox_max "10000"
- Limit max messages left in users inbox.
- quota_default
"1000mb" - Set reasonable per user default quota (this is a
domain level setting)
- In addition, be sure to use the 'config' checker in the web admin tool to
check recommended settings are being used.
In addition to the above you should consider the following
external monitoring for your system:
- Check for basic
TCPIP response on port 110,143,25
- Disk utilization
above 90% for 10 minutes.
- CPU utilization
above 50% for significant period of time (e.g. 30 minutes)
- Network utilization
above 50% for significant period of time (e.g. 30 minutes)
It would be wise to occasionally examine mail.err and
startstop.log manually for unusual issues.
Managers should also be aware of these
guidelines for disasters / critical problems.