Splunk Search

How to monitor transaction duration, but only account from 8AM to 6PM?

moisesroth
Path Finder

I have the fallowing search:

index="my_app" p_id=635392908992408562
| transaction p_id
| eval starttime=strftime(_time,"%Y-%m-%d %H:%M:%S")  
| eval endtime=if(duration!=0, strftime(_time+duration,"%Y-%m-%d %H:%M:%S"), "") 
| eval duration=if(duration!=0, tostring(duration,"duration"), "")
| table p_id starttime endtime duration

The result was:
alt text

Desired result:
In this case the duration should be 02:41:28, not 13:18:07, because the SLA should be account only between 8AM to 6PM.

Someone can help me to solve this issue?

Tags (2)
0 Karma
1 Solution

cmerriman
Super Champion

would something like this help as a starting point:

index="my_app" p_id=635392908992408562
 | transaction p_id
|eval starttime1=_time
|eval enddtime1=if(duration!=0,_time+duration, "")
|eval hours1=mvrange(starttime1,endtime1,3600)
|convert ctime(hours1) timeformat="%m/%d/%Y %H:%M:%S" 
|eval interestedHours=mvfilter(match(hours1,"( 08| 09| 10| 11| 12| 13| 14| 15| 16| 17| 18).*"))

that will get you a list of timestamps in hour intervals for every 8 am - 6 pm in your timeframe. you might want to do a eventstats maybe afterwards by p_id and date, perhaps, so that it resets every day, and then you can sum up the time frames each day. The caveat is that it wouldn't be exact H:M:S times, since it just adds a new timestamp every 3600 seconds.

View solution in original post

cmerriman
Super Champion

would something like this help as a starting point:

index="my_app" p_id=635392908992408562
 | transaction p_id
|eval starttime1=_time
|eval enddtime1=if(duration!=0,_time+duration, "")
|eval hours1=mvrange(starttime1,endtime1,3600)
|convert ctime(hours1) timeformat="%m/%d/%Y %H:%M:%S" 
|eval interestedHours=mvfilter(match(hours1,"( 08| 09| 10| 11| 12| 13| 14| 15| 16| 17| 18).*"))

that will get you a list of timestamps in hour intervals for every 8 am - 6 pm in your timeframe. you might want to do a eventstats maybe afterwards by p_id and date, perhaps, so that it resets every day, and then you can sum up the time frames each day. The caveat is that it wouldn't be exact H:M:S times, since it just adds a new timestamp every 3600 seconds.

moisesroth
Path Finder

Thanks! It works after some adaptations.
PS: There is an error at variable enddtime1, the correct should be endtime1

The final code for my environment was something like this:
One observation: Because I'm using minutes to count the total SLA, the regular expression used to validate from 08 to 18 should stop as 17, because 17 means 17:00, 17:01 ....17:59.

index="my_app" p_id=635392908992408562
| transaction p_id

| eval starttime=_time
| eval endtime=_time+duration
| eval my_minutes=mvrange(starttime,endtime,60)
| convert ctime(my_minutes) timeformat="%Y-%m-%d %H:%M:%S"
| eval interestedMinutes=mvfilter(match(my_minutes,"( 08| 09| 10| 11| 12| 13| 14| 15| 16| 17).*"))
| eval SLA_Minutes=mvcount(interestedMinutes)
| eval SLA_Time=tostring(SLA_Minutes*60,"duration") 

| eval h_starttime=strftime(starttime,"%Y-%m-%d %H:%M:%S") 
| eval h_endtime=if(duration!=0, strftime(starttime+duration,"%Y-%m-%d %H:%M:%S"), "") 
| eval h_duration=if(duration!=0, tostring(duration,"duration"), "") 

| table p_id h_starttime h_endtime SLA_Time
0 Karma

cmerriman
Super Champion

awesome, i'm glad it helped. i use a similar code for calculating weekdays v weekends, so i thought it could be similar for hours.

0 Karma

somesoni2
Revered Legend

What should be the duration if the starttime was 2017-05-15 14:00:00 with same endtime?

0 Karma

moisesroth
Path Finder
starttime 2017-05-15 14:00:00 and endtime 2017-05-16 10:41:28

The result should be 06:41:28.

0 Karma

somesoni2
Revered Legend

Getting complex 🙂 . What should be the duration if the starttime was 2017-05-14 14:00:00 with same endtime (stretched to more than a day)? Also, will there be cases where endtime be outside your SLA windows (falls between 6PM and 8 AM next day)?

0 Karma

moisesroth
Path Finder

1)
starttime 2017-05-14 14:00:00 and endtime 2017-05-16 10:41:28
The result should be 18:41:28.

On detail the calculation will be like below:
D14 - 14:00:00 to 18:00:00 = 04:00:00
D15 - 06:00:00 to 18:00:00 = 12:00:00
D16 - 06:00:00 to 10:41:28 = 02:41:28
SLA = 04h + 12h + 02:41:28 = 18:41:28

2)
Every starttime or endtime will be consider to the next valid time.
e.g:
19:00 will be consider next day at 08:00
06:00 will be consider current day at 08:00

0 Karma
Get Updates on the Splunk Community!

New in Observability Cloud - Explicit Bucket Histograms

Splunk introduces native support for histograms as a metric data type within Observability Cloud with Explicit ...

Updated Team Landing Page in Splunk Observability

We’re making some changes to the team landing page in Splunk Observability, based on your feedback. The ...

New! Splunk Observability Search Enhancements for Splunk APM Services/Traces and ...

Regardless of where you are in Splunk Observability, you can search for relevant APM targets including service ...