Extract and Transform Custom Event

jagadeeshm — Fri, 28 Apr 2017 11:21:02 GMT

I have events with the following format -

[Thread-2505_GOOGLE_INT_20170424155901301f9e61-1493049600619-NSRLM_2_1_RTL_39088504_2_R_PCLN,PCLN] 2017-04-24 12:00:02 : T:0.047 secs S:XXXX-SSSS-SSSSSS A:Availability M:INT_20170424155901301f9e61-1493049600619-NSRLM_2_1_RTL_39088504_2_R_1 CMD: [<?xml version="1.0" ?><AvailabilityRequestV2 xmlns="http://xml.google.com" siteid="1470249" apikey="SFGSDGSDFSDFGSFG" async="false" waittime="5"><Type>4</Type><Id>460573</Id><Radius>0</Radius><Latitude>0.0</Latitude><Longitude>0.0</Longitude><CheckIn>2017-10-01</CheckIn><CheckOut>2017-10-17</CheckOut><Rooms>1</Rooms><Adults>2</Adults><Children>0</Children><Language>en-us</Language><Currency>000</Currency></AvailabilityRequestV2>]

Notice couple of things - event starts with a thread name in square brackets, followed by date/time, followed by an unwanted ":", followed by several key/value pairs separated by a ":", and finally at the end you have content in XML inside square brackets.

I want to extract/transform this into the following so it is easy to search and run spath to extract fields from xml -

 Thread = Thread-2505_GOOGLE_INT_20170424155901301f9e61-1493049600619-NSRLM_2_1_RTL_39088504_2_R_PCLN,PCLN
DateTime = 2017-04-24 12:00:02
ResponseTime = 0.047 secs 
S = XXXX-SSSS-SSSSSS 
A = Availability 
M = INT_20170424155901301f9e61-1493049600619-NSRLM_2_1_RTL_39088504_2_R_1 
Action = CMD
XML = <?xml version="1.0" ?><AvailabilityRequestV2 xmlns="http://xml.google.com" siteid="1470249" apikey="SFGSDGSDFSDFGSFG" async="false" waittime="5"><Type>4</Type><Id>460573</Id><Radius>0</Radius><Latitude>0.0</Latitude><Longitude>0.0</Longitude><CheckIn>2017-10-01</CheckIn><CheckOut>2017-10-17</CheckOut><Rooms>1</Rooms><Adults>2</Adults><Children>0</Children><Language>en-us</Language><Currency>000</Currency></AvailabilityRequestV2>

Notice that the last key CMD is put into a new key "Action" and value is put inside XML. I was able to do some clean up like removing square brackets, removing extra ":" etc with the following SED Command -

SEDCMD-remove-xml-header = s/\<\?xml[^\>]*\>//g
SEDCMD-remove-square-brackets = s/\[|\]//g
SEDCMD-remove-colon = s/ : / /g
SECCMD-reame-threadname = s/Thread-/Thread:/1

How do I achieve the rest? Is a key value transformer?

Re: Extract and Transform Custom Event

jagadeeshm — Mon, 01 May 2017 22:49:21 GMT

Any splunk transformers 🙂 around?

Re: Extract and Transform Custom Event

beatus — Thu, 11 May 2017 14:26:26 GMT

Here's a regex that should work for you:
\[(?<Thread>.*?)\]\s(?<TimeStamp>(\d{4})-(\d{2})-(\d{2}) (\d{2}):(\d{2}):(\d{2})?)\s:\s(?:T:(?< ResponseTime>\S+\s+\S+))?(.)*(?<Action>(\w{3})):\s\[(?<XML>.*?)\]

This should handle the timeouts either existing in the log or not and only create the "timeout" field when they are there. Hope this helps!

Re: Extract and Transform Custom Event

jagadeeshm — Thu, 11 May 2017 14:29:05 GMT

Thanks beatus 🙂

topic Re: Extract and Transform Custom Event in Getting Data In

Extract and Transform Custom Event

Re: Extract and Transform Custom Event

Re: Extract and Transform Custom Event

Re: Extract and Transform Custom Event