Attempting to build some monitoring whereby we run a Splunk search from the command line interface (CLI) over a given timerange and that returns a certain count and then uses that count to trigger off alarms. I'm able to successfully get it all set up and the search is running successfully - but every time it runs, in addition to the count, it also returns "INFO: Your timerange was substituted based on your search string." Here's an example of the search we're running:
./search 'index=sandbox EventType=Content_Failed earliest=-15m@m |stats count'
This returns the INFO line above and then the count of "0" when things are good. The problem is that the INFO line is being interpreted by our third party system as part of the output of the search and ends up triggering off false alarms.
Is there any way to prevent that INFO line from appearing when running a CLI search with a specified timerange?
Instead of putting "earliest=15m" in your search, use the cli search options found here:
http://docs.splunk.com/Documentation/Splunk/6.5.0/SearchReference/CLIsearchsyntax
Example:
./splunk search 'index=_internal' -earliest_time -15m@m
Instead of putting "earliest=15m" in your search, use the cli search options found here:
http://docs.splunk.com/Documentation/Splunk/6.5.0/SearchReference/CLIsearchsyntax
Example:
./splunk search 'index=_internal' -earliest_time -15m@m
Completely forgot about this option. Good one!!!
Awesome - that fixed it. Thanks!
There doesn't seem to be an option to suppress the search warning/info messages in CLI. One workaround could be to 1) create saved search/report with same search string (without in-line time range) and specify the time range in saved search Start Time and End Time, 2) Change permission to be app level, 3) replace alert search by this
./search ' | savedsearch yourSavedSearchName' -app TheAppContextForSavedSearch
This also would have worked but unfortunately wasn't an option in our environment...