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

Is there anyway to exclude the Encrypter Data Exchange operations in DCRUM?

bcox2
Organizer

These operations are inflating the operation count of our applications and I would like to exclude globally. Is there a way to do that?

Thanks,

Brian

4 REPLIES 4

rodrigo_alvare1
Dynatrace Pro
Dynatrace Pro

My guess is to exclude it from your application definition and having it as a SS that you can consult to get information from SSL hanshake, etc.

Krzysztof_Ziemi
Dynatrace Pro
Dynatrace Pro

There is no way to decrease number of operations reported per server. These operations will always be counted per server:

  • for SSL decodes these will be reported under a specific name (to distinguish them from SSL connection setup)
  • for generic TCP decode, these will have no names, but would always be counted.

If the application-level reporting is the level on which you'd like to include only specific names in the counter (or exclude the others), then use the Business Units configuration to define operation name patterns per app or per transaction. This webinar https://university.dynatrace.com/education/dcrum/... shows how to do it for SSL connection setup operations, i.e.sepatate them out as a distinct transaction, so reports can be filtered to show URL calls and connection set-ups separately. Similar approach can be taken to encrypted data exchange operation names.

Best regards

Has anyone noticed a decrease in HTTPS URLs from version 12.4 to May2017? I am curious if some of the URLs are now reported in the "Encrypted data exchange" counts?

Krzysztof_Ziemi
Dynatrace Pro
Dynatrace Pro

Hi,

Separate bucket for encrypted operations wouldn't change the number of URLs recognized or number of page loads reported. However, compared to release 12.4 (classic AMD architecture), version 17 (HS AMD architecture) has many differences/fixes/enhancements in the HTTP decode that may change the number of URLs recognized and subsequently - number of pages loaded reported. HTTP decode has practically been re-built from ground up, so now it better covers many edge cases of page load sequence following.

It's possible to analyze specific cases in details, with the help of Support, and compare 12.4 and 17 measurements - but such analysis is time-consuming, so it really needs to be weighted against the benefits that such analysis result would bring.

Best regards