Microsoft SQL Server Extension 2.0 - Always On configuration - Duplicate Metrics

[36Krazyfists]

Hi,

We're setting up this extension and in the Hub page for it, regarding Always On monitoring, it states we should have two configurations for an Always On cluster:

- One configuration for all instances in the cluster (Primary and Secondary) but make sure the Always On feature set is turned Off.

- One configuration for just the primary replica instances in the cluster and ONLY the Always On feature set enabled.

Simple enough, but the issue is there is a "default" feature set that contains 13 metrics that you can not turn off.  This means that, for the Primary replica, it will have those 13 metrics duplicated (because it exists in the main config with all instances and in the Always On config).

Is this by design?  Or is there a better way to set this up?

Here is the docs and, if you scroll down to the FAQ, it has the section on setting up Always On clusters:

SQL Server Monitoring | Dynatrace Hub

I thought maybe the better way would be to just remove the Primary replicas from the non-Always On configuration and then, instead of enabling only the Always On feature set for that config, we would just turn all feature sets on.  This way we'd have one config for just the Secondary replicas with Always On unselected, and another config for only the Primary replicas with every feature set enabled including Always On.  No metric duplication of any kind this way.

However, I didn't know if maybe there was some logic requiring the Primary and Secondary replicas to all be in one config for it to properly recognize it as an Always On group or something.

Can anybody verify which method is appropriate (preferably somebody from the Dynatrace Extensions team or maybe somebody that has tried it both ways) ?

Also, maybe Dynatrace should update the documentation on the Hub page for this extension to better explain this, especially if having the two configs setup the way I suggested is preferable?

Thanks so much!

[AravindhanV]

If my understanding is correct, then we can use the tag to select the chosen Primary and Secondary cluster separated and create the configurations accordingly but Don't think we can choose the metrics to be captured since this is default provided by the product. And think we can serve this as a recommendation to create the "Feature Set". 

[zdstar]

@36Krazyfists I have found out the same problem. Did you sort it out somehow?
I see duplicate metrics -> duplicate problems -> duplicate notifications for metrics from default like user connections, blocking...
@AravindhanV I do not see way how to use tag for chosen primary because there is not any chosen primary. There can be done failover and then primary host will become secondary and secondary become primary host so it is impossible to set it statically.  What do you mean by "And think we can serve this as a recommendation to create the "Feature Set". " ?