I have the Splunk Add-on for Check Point OPSEC LEA installed on a number of instances and I'm a bit confused about what instance it is intended to be on.
I want the add-on for its ability to log into Check Point LEA, pull data, and send it to a destination index. As such, I want to use this add-on for inputs.
I have installed it on 3 heavy forwarders, none of which have indexes or run a search head, but the add-on appears to schedule an index=_* OR index=*
search every 10 minutes, even though it doesn't have a saved search. I think this search is part of a data model.
When I look in the $SPLUNK_HOME/var/run/splunk/dispatch//search.conf I can see
SearchParser - PARSING: summarize tstats=t override=partial manual_rebuilds=f max_time=3600 id=DM_Splunk_TA_opseclea_linux22_opsecMetrics [ search (index=* OR index=_*) (sourcetype=opsec) connection_name=* | eval nodename = "root" | rename connection_name AS root.connection_name | fields nodename, _time, host, source, sourcetype, root.connection_name ]
but I can't find the data model or saved search which is driving this.
Why is this enabled? I don't want it running on a heavy forwarder, so how do I disable this?
Disabling the data model is definitely in order if you don't need it to troubleshoot performance.
Is it possible that this app could be published in the form of separate inputs, parsing and input components ?
In our distributed environment, we run the inputs and parsing functions on splunk instances which do not have the web interface enabled.
For what it's worth - commenting the lines in default/datamodels.conf appears to have stopped this scheduled activity.