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

DEM Consumption reporting difference

mkulov
Organizer

Hi,

 

I am observing User sessions in total on the Environment details page (screenshot below) to be two time less that the amount reported by metric Session count - billed without replay.

Session replay is not enabled and there is only one application that consumes DEM.

Any idea why there is two times difference between both metrics?

Which one we should consider for license consumption?

 

mkulov_0-1645128639812.png

 

Thank you,

Martin Kulov

 

 

Digital Performance Optimizer
7 REPLIES 7

r_weber
DynaMight Champion
DynaMight Champion

Session count isn’t equal DEM units!

Depending on the type of session (real user, synthetic test,…) these consume 0.25 DEM units or 1 DEM unit. You need to do the maths yourself to go from sessions to DEM.

Certified Dynatrace Master, Dynatrace Partner - 360Performance.net

@r_weber I am comparing User sessions in first result and Session count in second result.

They should represent the same value, right?

 

Digital Performance Optimizer

r_weber
DynaMight Champion
DynaMight Champion

Hi @mkulov ,

 

what I experienced while building a central consumption monitoring is that you have to make sure to look at all billing metrics. (billing.synthetic, billing.apps.web, ...)

I usually exported those using my timeseries streamer and put them in a fine granular timeseries DB and then on a dashboard which allows me advanced queries/operations on that data. I found that this is pretty precise and always matched the consumption shown by Dynatrace on tenant-level.

Never noticed a difference, unless when not summing up with correct timeframes.

Screen Shot 2022-02-18 at 09.42.50.png

 

Did you consider the synthetic ones as well (though it shows 0 on your screenshot?) 

 

 

Certified Dynatrace Master, Dynatrace Partner - 360Performance.net

rastislav_danis
DynaMight Pro
DynaMight Pro

I did some tests and found this: If you display single value by SUM aggregation of metric "builtin:billing.apps.web.sessionsWithoutReplayByApplication" per actual month interval (since 1st day of actual month) you will get same number of sessions as number displayed in montly environment consumption.

 

If you have more apps then you have to count it for all of them:

builtin:billing.apps.web.sessionsWithoutReplayByApplication:filter(and(eq(Type,Billed))):splitBy(Type):sum

If you have also mobile ones, then +:

builtin:billing.apps.mobile.sessionsWithoutReplayByApplication:filter(and(eq(Type,Billed))):splitBy(Type):sum

Alanata a.s.

mkulov
Organizer

Thanks @rastislav_danis.

 

Yes, we are using SUM aggregation. 

However we are observing value twice bigger than the monthly license consumption.

 

Yellow box is twice bigger than the monthly black box in the first screenshot.

There is just one application consuming DEM and no session replay.

 

mkulov_0-1645171528495.png

 

 

 

 

Digital Performance Optimizer

But you forget to set timeframe to actual month (not last 30 days) !

Alanata a.s.

Yes, good catch.

The consumption is quite constant in this application though and cannot fluctuate by a factor of two.

Digital Performance Optimizer

Featured Posts