To add to that, the read intent only connections are handled via round robin method so if you have multiple readable secondaries, it will distribute them in that fashion to...

I like this topic. 

I might throw out there in addition to index / stats maintenance that you need to be checking for corruption regularly as well! Also, a...

I think that the endpoint would be fine. What I would be concerned about is having a synchronous replica on a remote site. Your round trip commits are going to...

+1 to SSRS recommendation. I have a collection of database / server level permissions scripts that I run quarterly for our audits manually. Haven't bothered with putting them in SSRS...

Have you manually ran the checkstep SQL in a query window on a secondary yet? I am really leaning towards that step not being good. It should properly raise an...

An error synchronizing should not affect whether the instance is considered a primary or a secondary under asynchronous mode.

This is amazing. I have created the proc and then created an agent job that executes this nightly on all of my replicas. Thank you!

If you run your check manually in a query window on a secondary, does it properly raise the error? I would start looking at the actual SQL at this point...

Are these app accounts or user accounts?

What does your step behavior look like? For example, after you run the is primary check in your first step, do you have to proceed to next step even on...

Logical reads / execution time 

What kind of index are you looking at? Do you have a way to simulate a production workload on your dev environment? My main...

If your servers are all sitting on the same rack (or under the same roof and not separated by region) and have the resources and IO, you could go with...

It obviously depends on your workload. How many queries do you want going parallel? 50 / 50? 

Here is a good blog post that provides a query that gives...