Process Item(s) / Trigger(s) faster than 1 second interval

[bryancan]

I have a table that is catching syslog data that is coming in from multiple hosts. This is a LOT of rows (or can be). I need to be able process these incomming rows. I think that I could make 1 item that has a lot of triggers that use regex to create events if a certain pattern is found.

How can I do this? There would be many rows per second so a 'check' at any interval would not work. I need to process every row as it is inserted into the syslog table...

Thanks,
Bryancan

[Alexei]

I would use monitoring of log files for this purpose. It is able to process much more than one single line per second.

[bryancan]

But where is the documentation for this ability?

[Alexei]

Look at the Manual, Tutorial for Log File Monitoring. While it does not give you very detailed information about all aspects of this functionality, it gives nice step-by-step instructions on how to configure monitoring of a log file.

[dmz]

Look at the Manual, Tutorial for Log File Monitoring. While it does not give you very detailed information about all aspects of this functionality, it gives nice step-by-step instructions on how to configure monitoring of a log file.

I've seen several references to this tutorial, but the only thing the tutorial shows is how to get the log file into zabbix. There is nothing there on actually monitoring the data.

I have the same question as the original post. How can I create a trigger that uses regex on the item it is checking?