I have a folder which contains multiple text files. I want to import these files into Splunk as events. which means each text file is 1 event. Each text file may have around 200 lines or less. Splunk seems to automatically split 1 text file to many events. What can I do to make Splunk recognize each file is an event?
*Note: each file could end with anythings, there is no specific ending pattern in the text file.
You can use BREAK_ONLY_BEFORE with an impossible string (well, relatively impossible) and it should look through the whole file without finding it and voila, you have a single file going into Splunk as a single event.
You can use BREAK_ONLY_BEFORE with an impossible string (well, relatively impossible) and it should look through the whole file without finding it and voila, you have a single file going into Splunk as a single event.
@cpetterborg: thank you so much. It's work. Is it ok if you post your answer below so I could accept it?
Here you go.
Does the text file start with a specific pattern?
You can play around with combination of below properties
BREAK_ONLY_BEFORE
TRUNCATE
SHOULD_LINEMERGE
http://docs.splunk.com/Documentation/Splunk/6.6.3/Admin/Propsconf
the file usually starts with the text File version. I don't know what could I do if it doesn't have a specific pattern at the end in which I thought we need that to do an event break.
All of these settings need to be on your indexer. It will make no difference if you put them on your search head.