I want to exclude part of JSON message before indexing. How can I achieve that> Below is a sample JSON. I used SED command in props.conf to exclude the first line and make it only JSON Indexed extraction. And another 2 more SEDs in props.conf to ignore a section of XML message (part of below JSON) as well
How can I write SED to ignore the "ignore" section below?
2019-10-02_09:09:09.234 My JSON message is
{A_message
{"blah": is blah
"blah2": is blah2
"blah3" : {
blah blah blah
}
"Ignore this" :
{
"ignore1": ignore1
ignore2: "ignore2"
}
}
Hi
You can use transforms to remove things before indexing.
[remove_ignore]
REGEX = (pattern before the data to ignore) data to ignore (pattern after data to ignore)
FORMAT = $1$2
DEST_KEY = _raw
Check the example in this answer:
https://answers.splunk.com/answers/485943/is-there-any-way-to-execute-sedcmd-after-transform.html
Hi nareshinsvu, this is a tough one. Until the data stream processor of Splunk is released, there is no easy way to efficiently do complex transformations on the data prior to ingestion. Depending on where the data comes from, you might want to take a look at implementing a modular input and preprocess the JSON in a small python script.