cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

How to calculate time between two purepaths

matt_lewis
Dynatrace Advisor
Dynatrace Advisor

Hey Community,

I have been told that this is not possible in AppMon (please correct me if wrong), so I was wondering if anyone has tried this before with an external tool?

The situation is we have a BT splitting on transaction ID. This results in two purepaths:

  1. Query Place Request (/b2b)
  2. Query Place Response (DynatraceWorkerB2B....)

The Response only comes through after the Request purepath has completed. What I would like to do is calculate the time between the two start times. I can use the BT to drill down to the Request/Response pair and calculate the difference manually (difference between the two start times, plus the Response time of the Response Purepath), but I would ideally like to lay this out in a chart, so we can see total time for these two to complete.

Any ideas or thoughts on how this is accomplished (inside of, or outside of DT) would be much appreciated.

Thanks,

Matt

3 REPLIES 3

Hi Matt,

I think you can create a dashboard that has your BT on it and then use a RESTful server side scripting platform to bring up the XML version of your dashboard (i.e. https://localhost:8021/rest/management/reports/cre... , collect the data that you need (start times), do the calculations, and then report on it.

Hope this helps,
Dan

Hey Dan,

Thanks for the tip, will be sure to give this a try!

-Matt

Hello Matt,

We also did rest integrations to retrieve purepaths, you can find info on how to do that in below forum link: https://answers.dynatrace.com/questions/164930/usi...

However there is an important remark related to your question,

The rest API has been designed to manage your AppMon environment. Getting complete purepaths by using the rest is possible, however you need to do it with caution. High throughputs for this kind of request can bring your server health at risk. Even low throughputs with high time periods into the request can be a risk, definitely on production environments. So the question is how many of those purepaths do you need to retrieve and how big are those pureppaths (#,nodes)?

There are other solutions to stream data to external systems, such as elastic search, splunk, InfluxDB (timeseries), ... by using the purelytics (POST/elastic) stream or the powerful business transaction feed. Then you can perform your analysis on elastic for example.

Regards