I would like to get more detail on an error message in our Dynatrace Activegate logs. In particular this error message:
2025-08-01 18:11:19 UTC INFO [<DYN-TENANT-ID>] [RestModel] Socket closed by Observable. Throttling error for the 1 attempt out of maximum 8 for service VT <SUBSCRIPTION-ID>/ResourceGraphProvider/Generic Entity . Retrying after delay of 3259ms.
There is a 12,000 metric limit per subscription per hour through Azure Monitor. I do not believe we are hitting this 12K limit per subscription. One item that goes to that fact is that the Dynatrace doc says that the polling is set to every 5 minutes (300 seconds) for the Azure Monitor integration. If Dynatrace calculates that the polling will hit the 12K limit per hour per subscription, it will back off and change the polling to less frequent intervals (per doc here).
Most, if not all, of our subscriptions do NOT have enough resources to push this limit and I have NOT seen any polling interval more than 300 seconds in the dynatracegateway_Debug.*.log files.
We are trying to come up with a strategy that will keep the admin overheard down, but still ingest reliable data. Any insight into this would be appreciated.
Thanks!
k.
