We are reporting events in a CSV format using a comma deliminated structure. Example below:
,playerID04,player01,client002,clip_started,false,active,,237
Essentially this is a list of info sent when an event happens, comprised of:
player ID, player type, client ID, the event type (clip_started), parameter1, parameter2, video ID (237 in this instance).
In the 1st instance it is the video ID that I am after
What I want to produce is a list or table of these video IDs, showing how many times they show, or are associated with the 'clip_started'. The video ID isn't named as such, but it always appears in the same position within the data, ie. after the 8th comma. Is there a way for me to tell splunk that this is a VIDEO-ID, to extract this number in a repeatable way?
To further expand on this, is there a way for em to tell splunk the names of all of these fields, and for splunk to automatically 'name' them as fields that I can then filter and search by.
thanks
Comma delimited file means the CSV files which splunk by default recognizes. Do you have headers on the files? Have you used the inputcsv or inputlookup command to read the content and filter them afterwards?
Yes. Setup a delimiter based field extraction.
In props.conf, you'd do something like
[yoursourcetype]
REPORT-video = videoinfo
And in transforms.conf
[videoinfo]
DELIMS = ","
FIELDS = "player_ID", "player_type", "client_ID", "event_type", "parameter1", "parameter2", "video_ID"
looking. thanks
Yes. Have a look at the extract command. http://docs.splunk.com/Documentation/Splunk/latest/SearchReference/Extract
Is it possible to do this from the search area?