Splunk Search

Regex Help for special characters

chanukhya
Explorer

Hi,

My log looks like this. I am trying to get the average response time by service.

ServiceInvoker (service_A) : executeFlow : Time Take is = 3378
ServiceInvoker (service_B) : executeFlow : Time Take is = 378
ServiceInvoker (service_C) : executeFlow : Time Take is = 338

Here is what i have:

 index=app  |rex '\ServiceInvoker\s+"((?<service>\S+))"\s+:\s+executeFlow\s+:\s+Time\s+take\s+is\s+=\s+(?<response_time>\d+)'   | stats  sparkline(avg(response_time),1m) as processTime_trend, avg(response_time),count BY service 

The brackets that are surrounding the service name is causing an issue for retrieving the results. Any help or ideas would be appreciated.
Thanks in advance

0 Karma
1 Solution

gokadroid
Motivator

Try this:

index=app 
|rex field=_raw "(.*|^)ServiceInvoker\s*\((?<service_name>[^\)]+)\)\s*.*Time\s*Take\s*is\s*\=\s*(?<respTime>[\d]+)"
| stats  sparkline(avg(respTime),1m) as processTime_trend, avg(respTime),count BY service_name

See here the regex in action

View solution in original post

lakromani
Builder

Do you need all the line in the regex? If not, you can do like this:

index=app  | rex "Time Take is =\s(?<respnse_time>\d+)"  | stats  sparkline(avg(response_time),1m) as processTime_trend, avg(response_time),count BY service 
0 Karma

gokadroid
Motivator

Try this:

index=app 
|rex field=_raw "(.*|^)ServiceInvoker\s*\((?<service_name>[^\)]+)\)\s*.*Time\s*Take\s*is\s*\=\s*(?<respTime>[\d]+)"
| stats  sparkline(avg(respTime),1m) as processTime_trend, avg(respTime),count BY service_name

See here the regex in action

aljohnson_splun
Splunk Employee
Splunk Employee

Hi @Chanukhya,

Just escape the ( with a backslash.

https://regex101.com/r/BTBkvw/1

ServiceInvoker\s+\((?<service_name>\w+)\)\s+:\s+(?<service_flow>\w+)\s+:[^=]+=\s(?<response_time>\d+)
0 Karma

chanukhya
Explorer

Sorry, It didn't work. Updated my question.

0 Karma

sundareshr
Legend

Try this

*UPDATED

index=app  |rex "[^\(]+\((?<servicename>[^\)]+)\)[^=]+=[\s\t]+(?<response_time>\d+)"  | stats  sparkline(avg(response_time),1m) as processTime_trend, avg(response_time),count BY service

*OR*

index=app  |rex "[^\(]+\((?<servicename>[^\)]+)" | rex "=[\s\t]+(?<response_time>\d+)"  | stats  sparkline(avg(response_time),1m) as processTime_trend, avg(response_time),count BY service

chanukhya
Explorer

Thanks for your help.

0 Karma

gcusello
SplunkTrust
SplunkTrust

Remember to put backslash before =
Bye.
Giuseppe

0 Karma

chanukhya
Explorer

Sorry, It didn't worked.

0 Karma

sundareshr
Legend

Try the updated query

0 Karma

chanukhya
Explorer

It didn't worked as well, The service names are different and some service names has an underscore in the name and some dont. I am trying to get the average response times and count for each service, which is in between the brackets.

0 Karma

sundareshr
Legend

Do you see any results when you try this

index=app  | rex "[^\(]+\((?<servicename>[^\)]+)\)[^=]+=[\s\t]+(?<response_time>\d+)"  | table servicename response_time
0 Karma
Got questions? Get answers!

Join the Splunk Community Slack to learn, troubleshoot, and make connections with fellow Splunk practitioners in real time!

Meet up IRL or virtually!

Join Splunk User Groups to connect and learn in-person by region or remotely by topic or industry.

Get Updates on the Splunk Community!

Improve Delivery Assurance with S2S ACK for Edge Processor

Edge Processor helps Splunk customers process data closer to the source: filtering, transforming, masking, and ...

.conf26 Platform Sessions: Turn Machine Data into Agentic Action

As autonomous agents and multi-cloud architectures reshape modern IT, data platforms have to do far more than ...

Introducing the Launch of Edge Processor in Hybrid Mode!

Modernize Data Ingestion Without Starting Over  For years, organizations have relied on Splunk's proven ...