On premise data gateway monitoring

[al_cslai]

Hi, 

Have someone already tried this template?

https://learn.microsoft.com/en-us/data-integration/gateway/service-gateway-performance

Would love to understand how the refresh works and how you can schedule it so there's a report that we can monitor

[lbendlin]

Don't waste your time. That template is very old, only works with a single gateway cluster member, and is missing all the new logs.   Create your own monitoring process. Focus more on real time performance and keep the logs only for the occasional forensics.

[al_cslai]

He.. seems typical 😅 Any tips or tricks for creating own monitoring process? Is it simply just using the Event Viewer or Performance Monitoring. Any other sort of metrics that needs to be checked?

[lbendlin · Accepted solution]

In my opinion these are the things that need to be tracked on each gateway cluster member:

- Proxy availability and network delay  - no proxy = dead gateway

- processor utilization across cluster members. 80% and below is ok.  Make sure nothing else is running on these VMS

- memory utilization ( but that you can and should limit to 95% in the config file)

- number of mashup containers

We do this via Powershell scripts on each of the cluster members, and the telemetry is then pushed into a hybrid streaming dataset.

Here's an example from one of our gateways