Splunk Search

How to calculate the percentage increase (or decrease) in each error/success code based on previous hour?

onthakur
Explorer

Team, 

I have below timechart which is counting http error/success codes for a span of 1hr.

Now I need to calculate the percentage increase (or decrease) in each error/success code based on previous hour.

_time 200 4xx errors 5xx errors
2022-05-23 00:00 100 20 30
2022-05-23 01:00 200 30 30
2022-05-23 02:00 250 50 60
2022-05-23 03:00 300 30 50
2022-05-23 04:00 350 40 40
2022-05-23 05:00 400 60 60
2022-05-23 06:00 500 80 80
Labels (1)
0 Karma

somesoni2
Revered Legend

Try this generic search (will work with any status)

Your current search with timechart
| streamstats current=f window=1 values(*) as prev_*
| foreach prev_* [| eval "PercChange_<<MATCHSTR>>" =round(('<<MATCHSTR>>'-'<<FIELD>>')*100/'<<FIELD>>',2)]
0 Karma

ITWhisperer
SplunkTrust
SplunkTrust

Try something like this

| streamstats current=f window=1 latest('200') as previous_200 latest('4xx errors') as previous_4xx_errors latest('5xx errors') as previous_5xx_errors
| eval 200increase = 100 * '200' / previous_200 - 100
| eval 4xxincrease = 100 * '4xx errors' / previous_4xx_errors - 100
| eval 5xxincrease = 100 * '5xx errors' / previous_5xx_errors - 100
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!

Splunk App Dev Quarterly Roundup: AI, Agents, and Innovation!

Another quarter, another wave of innovation. From complex integrations to pushing the limits ...

What’s New in Splunk AI: Volume 02

Welcome to the second edition of “What’s New in Splunk AI” where we look at the latest and greatest updates, ...

Value Insights: Now Generally Available in the CMC

Organizations are under pressure to move faster, control cost, expand AI adoption, and prove value with more ...