Getting Data In

trigger a script on a Forwarder from Indexer on another host?

Path Finder

I've got a single v4 Splunk Indexer/Search. Feeding it are multiple Forwarders that have local indexing disabled and have all their inputs/conf deployed to them in apps. I'd like for a scheduled search on the Indexer to be able to trigger a script I've deployed on the forwarder.

My limitations are that I don't want the script to run and send data continuously (which would be easy) because 99% of the time it's not relevant, and we're trying to isolate a problem that occurs sporadically. Also, our environment doesn't allow password-less (automated) logins from host-to-host, so the simple ssh solution isn't available.

There's doesn't appear to be anything about this in help/answers, but I may be searching for the wrong thing. Can Splunk do this? Am I going about this the wrong way? Thanks.

Tags (3)
0 Karma
1 Solution

Splunk Employee
Splunk Employee

Hm, Splunk is not generally designed to have such a channel.

You could perhaps abuse the distributed search functionality to cause a distributed search to trigger a 'command' implemented by a script, which actually updates a log which the nodes monitor. Or something like that. It's almost certainly more work than you'd want, and quite rube-goldberg.

Personally I'd have the input scripts hit a url and if they get a successful response of say, an integer, and that integer is updated, produce the output. Then you can fairly easily update such a number on an http of your choosing or by modifying a static asset in a splunkweb directory.

Ie, personally I'd make the trigger mechanism be external to Splunk.

View solution in original post

0 Karma

Splunk Employee
Splunk Employee

Hm, Splunk is not generally designed to have such a channel.

You could perhaps abuse the distributed search functionality to cause a distributed search to trigger a 'command' implemented by a script, which actually updates a log which the nodes monitor. Or something like that. It's almost certainly more work than you'd want, and quite rube-goldberg.

Personally I'd have the input scripts hit a url and if they get a successful response of say, an integer, and that integer is updated, produce the output. Then you can fairly easily update such a number on an http of your choosing or by modifying a static asset in a splunkweb directory.

Ie, personally I'd make the trigger mechanism be external to Splunk.

View solution in original post

0 Karma

Path Finder

Thanks, that's a good idea given my restrictions. I have a common NAS share that should serve as a means of communication between the two pieces, and I'm already thinking of other uses for "conditional scripts".

0 Karma