Splunk Search

Regex with conditional statement

krusovice
Path Finder

Hi there,

I need some help to form regex command. My requirement is to first search for code=SEND then stats count the CountryCode, secondly search for code=RECEIVE then stats count the CountryCode.

This is my XML log:

                    <Cust>
                    <Code>SEND</Code>
                           <CountryCode>CN</CountryCode>
                           <Lty>
                              <CtyNm>BEIJING</CtyNm>
                              <Zip>100176</Zip>
                           </Lty>
                  </Cust>
                  <Cust>
                     <Code>RECEIVE</Code>
                           <CountryCode>JP</CountryCode>
                           <Lty>
                              <CtyNm>TOKYO</CtyNm>
                              <Zip>1000001</Zip>
                           </Lty>
                  </Cust>

I'm having this query formed but not meeting the above requirement, it only matches the first code=SEND and perform the count.

index=*
| rex "(?P[\w+\s+]+)"
| stats count by country

Appreciate for your help.

Tags (2)
0 Karma
1 Solution

493669
Super Champion

If you do the ingestion in such a way that each is a separate event in Splunk with valid xml syntax, the field extraction can be done by adding KV_MODE = xml in props.conf on Search Head(s). So, main focus should be getting the data ingested correctly. if your data doesn't have timestamp then here current time is considered as the _time value for the event.

Try this for props.conf on your Indexer/Heavy Forwarder.

 [SourceType_name]
 SHOULD_LINEMERGE = false
 LINE_BREAKER = ([\r\n]+)(?=\s*\<Cust\>)
 DATETIME_CONFIG = CURRENT

Search head props.conf

 [SourceType_name]
 KV_MODE =xml

If you separate events on the basis of tag then it will be easier to perform evaluation

View solution in original post

0 Karma

493669
Super Champion

If you do the ingestion in such a way that each is a separate event in Splunk with valid xml syntax, the field extraction can be done by adding KV_MODE = xml in props.conf on Search Head(s). So, main focus should be getting the data ingested correctly. if your data doesn't have timestamp then here current time is considered as the _time value for the event.

Try this for props.conf on your Indexer/Heavy Forwarder.

 [SourceType_name]
 SHOULD_LINEMERGE = false
 LINE_BREAKER = ([\r\n]+)(?=\s*\<Cust\>)
 DATETIME_CONFIG = CURRENT

Search head props.conf

 [SourceType_name]
 KV_MODE =xml

If you separate events on the basis of tag then it will be easier to perform evaluation

View solution in original post

0 Karma

krusovice
Path Finder

I'm not familiar with props.conf, the ... is nested inside a long output of XML, and there are many events with these tags. Could you further explain by editing props.conf, how can I achieve what I needed to query?

0 Karma

493669
Super Champion

have you indexed your xml log ?
if not then you can separate events on the basis of start/end of tag
like here in your case I supposed your starting tag is <Cust> so on the the basis of this I have separated events .
refer LINE_BREAKER = part in docs
http://docs.splunk.com/Documentation/Splunk/latest/Admin/Propsconf?utm_source=answers&utm_medium=in-...

0 Karma

krusovice
Path Finder

Thanks for explaining and I got your point now. The log already indexed and I can't un-do it since I'm not admin. Furthermore, the log is not purely just XML but with other text and SOAP messages as well.

0 Karma

493669
Super Champion

try this:

   ... |rex field=_raw max_match=0 "(?m)<code>(?<code>[^<]+).*(?[^<]+)"|eval merged=mvzip(Code,CountryCode,";;")|mvexpand merged| eval merged=split(merged,";;") | eval Code=mvindex(merged,0)| eval CountryCode=mvindex(merged,1)

try this run anywhere search:

|makeresults|eval raw="<Cust> <Code>SEND</Code> <CountryCode>CN</CountryCode> <Lty> <CtyNm>BEIJING</CtyNm> <Zip>100176</Zip> </Lty> </Cust>
<Cust> <Code>RECEIVE</Code> <CountryCode>JP</CountryCode> <Lty> <CtyNm>TOKYO</CtyNm> <Zip>1000001</Zip> </Lty> </Cust>"
|rex field=raw max_match=0 "(?m)<Code>(?<Code>[^<]+).*<CountryCode>(?<CountryCode>[^<]+)"|eval merged=mvzip(Code,CountryCode,";;")|mvexpand merged| eval merged=split(merged,";;") | eval Code=mvindex(merged,0)| eval CountryCode=mvindex(merged,1)
0 Karma

krusovice
Path Finder

Thank you. I get some ideas with your answer.

0 Karma

493669
Super Champion

is this only one event? why you have not separated event on basis of tag

0 Karma
Register for .conf21 Now! Go Vegas or Go Virtual!

How will you .conf21? You decide! Go in-person in Las Vegas, 10/18-10/21, or go online with .conf21 Virtual, 10/19-10/20.