Azure Virtual Desktop Round Trip Time

[StoleSursum]

Reaching out to ask if anyone has suggestions for pulling the round-trip times for each AVD machine.  This information is available local to each AVD machine (reference the attached screenshot).  However, I need this metric ingested for all AVD machines and pulled to a dashboard showing the high latency times (or not) as it relates to 3000+ machines concurrent.  

It is is in relation to providing an option for Tier 1 and 2 teams to troubleshoot high number of RDP disconnects and slowness complaints.  The metrics available in DT and with the OneAgent installed in all AVD is not what we need.  Actually, those network metrics are useless as it relates to RTT, troubleshooting RDP disconnects, and measuring latency of RDP sessions.

According to vendor documentation, it is possible to ingest WMI data leveraging our current implementation of OneAgent and enable a separate DT component Extension Execution Controller.

EEC and WMI run alongside Windows OneAgent locally within a VDI, and no other requirements. ActiveGate was excluded from the discussion based on our requirements and leveraging EEC with WMI extension alongside Windows OneAgent.

Reference: https://www.dynatrace.com/support/help/extend-dynatrace/extend-metrics/ingestion-methods/wmi

https://www.dynatrace.com/support/help/shortlink/extend-wmi#before-you-begin

Caveats related to additional resource consumption in VDI with EEC enabled.  Basically, enabling this functionality might consume more resources VDI side. 

[Reference: https://www.dynatrace.com/support/help/shortlink/extensions-concepts#resource-consumption]

Curious if anyone else has accomplished this for AVD machines where a lot of the information needed to troubleshoot is available from within a single session on per session basis and we need that in Dynatrace and fitted to a Dashboard that support teams can actually use to troubleshoot problems and possibly predictive analysis once we have the metric data.

Or can point me in a direction and I'll take it from there.

[ChadTurner]

@StoleSursum were you able to pull that data in via  the Data explorer then split and filter on the particular dimensions that best suit your use case?