You could create a script that computes the counts you're looking for, a naïve one-liner counting the files and directories in a path could be this:
cd /path/to/count; date; pwd; ls | wc -l
Then configure that as a scripted input in your Splunk.
I'm sure similar things can be done with PowerShell. A quick googling suggests something like this: http://stackoverflow.com/questions/14714284/count-items-in-a-folder-with-powershell
Thank you martin, but i'm working with windows not linux, so what should i do ?
Sure. Say you saved my one-liner or your much better script to $SPLUNK_HOME/etc/apps/your_app/bin/dircount.sh
, then you can define this in inputs.conf:
[script://$SPLUNK_HOME/etc/apps/your_app/bin/dircount.sh]
index = your_index
sourcetype = your_sourcetype
interval = number of seconds between runs
You can also configure this through the UI if run on a Searchhead, Indexer, or Heavy Forwarder by going to Settings -> Data Inputs -> Scripts -> New.
Find the docs here: http://docs.splunk.com/Documentation/Splunk/6.1.3/Data/Setupcustominputs
Can you expand on the "Then configure that as a scripted input in your Splunk." Thanks!
I want just to calculate the number of directories and files in a specific folder ! so if it is possible to not index data it will be good. If not i want to have the count and create a dashboard like that :
DIRECTORY NumberOfFiles NumberOfDirectories
Dir Name NBFiles NBDir
So How could i do this ?
Regards
Are you trying to index that count for some directory, or are you trying to calculate statistics based on already indexed data and the path in their source
field?