Here goes. I use snmp to keep an eye on the acceptable values for loadavg since I systematically use cricket
and nagios to graph such stuff on all systems I'm supporting.
----------------------------------------------------------------------
#!/bin/sh
MAILTO="[EMAIL PROTECTED]"
WHEREAMI=`hostname`
get_snmp() {
/usr/bin/snmpget -Ovq -v 1 -c public localhost "$1"
}
send_alert() {
mail -s "[${WHEREAMI}:$1]" $MAILTO </dev/null
}
check_load() {
# trigger error when 15min too high
# and neither 1min nor 5min below threshold
avg=`get_snmp "laErrorFlag.1"`
[ "$avg" = "0" ] && return 0
avg=`get_snmp "laErrorFlag.2"`
[ "$avg" = "0" ] && return 0
avg=`get_snmp "laErrorFlag.3"`
return $avg
}
restart_dbmail() {
send_alert "DbmailRestartAlert"
/etc/init.d/dbmail restart
}
check_load || restart_dbmail
-----------------------------------------------------------------
ocl wrote:
I do run 2.0.1 in production, but not without running a cronjob which
checks the load on a machine, and restarts dbmail when the load
becomes excessive.
Could you please post a copy of this cron script.
It sounds like a good idea for all seasons.
Cheers,
Ray
_______________________________________________
Dbmail mailing list
[email protected]
https://mailman.fastxs.nl/mailman/listinfo/dbmail
--
________________________________________________________________
Paul Stevens mailto:[EMAIL PROTECTED]
NET FACILITIES GROUP PGP: finger [EMAIL PROTECTED]
The Netherlands________________________________http://www.nfg.nl
