Creating a new field based on substring

[Pradana]

I have below message :
2018-06-08 21:35:44,182 DEBUG [org.mobicents.smsc.library.CdrGenerator] 2018-06-08 21:35:44.171,62895416500343,1,1,6282880640846,1,1,success_esme,SS7_HR,message,null,92044,0,null,null,null,null,628964011092,null,0,15,null,0,0,,,,2,"Yth.Bpk/Ibu kredit d","",,,

I filtered it using :
%{TIMESTAMP_ISO8601:logdate} %{LOGLEVEL:debugtype} %{DATA:source} %{TIMESTAMP_ISO8601:smsdate},%{WORD:sourceaddr},%{NUMBER:addrton},%{NUMBER:addrnpi},%{WORD:destaddr},%{NUMBER:sourceton},%{NUMBER:sourcenpi},%{WORD:status}

I want to add another filed, let says "operator" based on the first 5 digit of sourceaddr.

How i can achieved above objective ?

[Badger]

This provides an example of extracting a fixed number of characters as a prefix to a field.

[Pradana]

Thanks for your comment. I have read the sample but still i am not sure how to apply it. Can you elaborate more on this ?

[Christian_Dahlqvist]

Have you tried a separate grok block with the following:

grok { match => [ "sourceaddr", "^(?<operator>.....)" ] }

[Pradana]

I try like below in https://grokdebug.herokuapp.com/ :

%{TIMESTAMP_ISO8601:logdate} %{LOGLEVEL:debugtype} %{DATA:source} %{TIMESTAMP_ISO8601:smsdate},%{WORD:sourceaddr},%{NUMBER:addrton},%{NUMBER:addrnpi},%{WORD:destaddr},%{NUMBER:sourceton},%{NUMBER:sourcenpi},%{WORD:status}[ "sourceaddr", "^(?.....)" ]

Above pattern doesn't give me another field operator.

[Christian_Dahlqvist]

Put it in two separate grok blocks. The first one extracts all the fields and the second the 5 character prefix.

[Pradana]

Hi Christian,

Thanks for your suggestion. Its working. Now I have a field call operator with only 5 string on it. But I have another problem, it seems, after adding this field, the field is not part of the index. How I include them into index ?

[Pradana]

Hi,

It seems the issue is in Kibana. After adding the field to become index in kibana, the field can be visualized now.

Thanks