|
I try to add some csv files, which contain data like the followings Time, ACTION,ORDER_NO, ... 2009-11-2 20:00:00.041,REQUEST,48613840, ... 2009-11-2 20:00:00.041,REQUEST,48613839, ... 2009-11-2 20:00:00.041,REQUEST_ACK,48613840, ... 2009-11-2 20:00:00.041,REQUEST_ACK,48613839, ... 2009-11-2 20:00:00.046,REQUEST,48613841, ... when I set the input source type as "csv", then the input file can be recognized with an "AutoHeader-1" stanza and a "csv-2" stanza being added to "$Splunk\etc\apps\learned\local\transforms.conf" and "$Splunk\etc\apps\learned\local\props.conf" respectively. But I still have two problem,
|
|
in your props.conf also add this line: [yoursourcetype] TRANSFORMS-NoHeader = NoHeader on your transforms.conf add this: [NoHeader] REGEX = Time, ACTION,ORDER_NO, ... DEST_KEY = queue FORMAT = nullQueue |
|
|
Seriously folks, the solution in answer <1> should be the default behaviour. Why would you want the headers included in your indexed data??? Do you really want Splunk choosing (own its own) to just drop certain events out of your log files?
(02 Aug '11, 10:43)
Lowell ♦
|
