I have a log file that my custom application uses named 'application.log'.
I want to set up a daemon that will monitor this file and scan for any errors marked CRIT and ignore the other DEBUG and WARN messages. The daemon should be easily configurable to work with other applications, so that the logging mechanism can be standardised across the network.
Furthermore, the software should be mature in the sense that it should have an extensive feature set. For example, the software should be intelligent enough to send multiple errors occurring within 30 seconds in one email rather than spamming duplicate notification events.
I was told that syslogd can be used to this effect? I am currently diving blindly into the documentation, but I would very much appreciate some guidance from an experienced sysadmin. ;-)