Splunk Search

How to group and count data from string?

vintik
Engager

I have the following query:

sourcetype=XXX Some query for * took * seconds to load

And this is a result of query:

I, [2016-09-28T11:01:21.616144 #23942]  INFO -- : Some query for authorizations took 8 seconds to load. JobID: 1475060473.4330475
I, [2016-09-28T11:01:21.113586 #20705]  INFO -- : Some query for authentications took 9 seconds to load. JobID: 1475060472.4330474
I, [2016-09-28T11:01:20.539219 #19285]  INFO -- : Some query for authentications took 8 seconds to load. JobID: 1475060472.4330473

But I need to get average time for each of operation type (authorizations and authentications from example):

authorizations: 8.5
authentications: 8

Is it possible?

0 Karma
1 Solution

somesoni2
Revered Legend

Assuming that both operation type and duration is not extracted, try like this

 sourcetype=XXX Some query for * took * seconds to load | rex "for (?<operation_type>\w+) took (?<duration>\d+) seconds" | stats avg(duration) as Average_Time by operation_type

View solution in original post

somesoni2
Revered Legend

Assuming that both operation type and duration is not extracted, try like this

 sourcetype=XXX Some query for * took * seconds to load | rex "for (?<operation_type>\w+) took (?<duration>\d+) seconds" | stats avg(duration) as Average_Time by operation_type
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!

Catalog Is Now Generally Available on Splunk Cloud Platform

A Unified View of Your Data  Security logs, application events, business data, and historical telemetry often ...

Your Path to AgenticOps: AI Experiences for Every Splunk Practitioner

Your Path to AgenticOps: AI Experiences for Every Splunk Practitioner   Watch on demand us for a demo-driven ...

The Data Connection - October 2026

We all have data. Lots of it. But that’s not the problem – it’s understanding what we should we do with it, ...