Hello,
We have UEM implemented with a pretty basic configuration to keep the overhead low. Now the customer wants to enable the Perceived Render Time metric, but in order to do that we need to know what additional overhead will be generated by enabling this feature.
Is there a way to calculate the additional overhead to the dT server, if any?
Thanks!
Answer by Joe H. ·
Most of the overhead related to this metric is in the collection and calculation of the data on the browser, not in the dTServer or collector itself. Simply to collect the metric, should not have a measurable change to the load on the Collector or dT Server. Adding BTs as a result of this metric is a more likely place where you'll create additional overhead on the dTServer, so just be reasonable and you'll likley not notice any change.
Answer by Luciana S. ·
Thanks Andi.
I've created a script using Gomez and added dT tag to facilitate tracking. I couldn't see any increase in the response time or purepath size. I also checked dtServer health + collector sizing and it looks the same as before.
Based on that, would it be safe to assume that activating Perceived Render Time will not cause any significant overhead to the dtServer or to the collectors?
Right now their main concern is regarding dTServer and collectors because they are running at maximum capacity.
Answer by Andreas G. ·
Hi
There is no way to calculate the overhead. In order to calculate Perceived Render Time our JavaScript agent has to iterate through the DOM to figure out whether the initial page is fully loaded. Depending on the actual web page (# of DOM Elements), browser (type + version), the JavaScript Engine and the processing power of the machine this can vary. Therefore it is hard to tell.
Best is to test this out by e.g: using a synthetic script to measure the execution time of a specific set of steps with and without Perceived Render Time.
Andi
JANUARY 15, 3:00 PM GMT / 10:00 AM ET