On 03/11/04 19:32 -0800, Trevor Warren wrote: <snip>
In my network consisting of purely AIX/Gnu-Linux/HP-UX i have to provide a single console for monitoring all logs and trigger an alert/alarm as soon as a particular message comes up in the logs. These logs will be both system and application logs and hence the messages to be tracked has to be customisable.
Log to syslog, over TCP. Centralized syslog server. Have a Perl script watch for those error message patterns. Have the script send out a snmp trap. Trivial to implement.
Devdas Bhagat