SQL MP wrongly discovers Service Pack Level attribute which shows 0 but the SQL 2016 is already installed as 13.1 which is SP1 version. Please correct.

To exclude an instance or database from discovery, you have to either specify the full name or use * to exclude everything. It would be much more use to be able to exclude '*_clone' or '*test*'.

SCOM is apparently assuming that a machine with SSMS (SQL Server Management Studio) installed must also have SQL Server installed and is alerting with "MSSQL 2014: Monitoring warning" or "MSSQL 2016: Monitoring warning" for all of our RDS VMs used by our IT group. SSMS has been a separate install for quite some time now and the SQL Management Pack needs to be updated to handle that properly.

Today the SQL MP will not check the IP adresse the SQL Instance is listen. The SQL MP looks for the port. It then assumes that the IP Adresse is the adresse of the FQDN of the server. But this is not always the case. Please improve the system and look for the IP adresse. Then you can derive the right FQDN.
In my opinion at the moment this is just not correct implemented.

Thank's

Please add tasks to Stop/Start SQL 2014/2016 Analysis Services and Reporting Services. This is available in SQL 2005-2012 MP but was not put into the SQL 2014-2016 SSAS/SSRS MP's. SQL 2014-2016 DB Engine has these tasks. We have DBAs that have used this task in SCOM for years but are unable to Stop/Start services since SQL 2014.

The scripts in SQL 2014 MP does not handle AlwaysOn correct. When evaluated if the database is a replica the property secondary_role_allow_connections is checked and if set to <= 1 the database is skipped. However, this is true even if the script is connected to the primary node which is accepting connections. This causes all free space monitoring for AlwaysOn databases that do not allow connections to secondary nodes to fail without any notice to the users, even when the script is connected to the primary node.

Currenly it is not possible to configure the SQL Agent Job monitors Long Running Jobs and Last Run State to check on specific days a week. The parameters Start Time, End Time and Days Of Week Mask would be most welcome.

Many SQL Agent jobs are configured to run once week or even once a month.

SQL Datacenter Dashboard does not display additional Performance/Monitor widgets added on initial display. It only displays the additional widgets if I attempt to add an additional one. I would expect all of the widgets selected including the group, performance, and monitor pieces to display everytime I visit the dashboard. (Using the widget to build a dashboard for Hyper-V and Active Directory groups of objects.)

VBScripts that log into the SQL instance attempt to open secondary copies that are (intentionally) set to not be readable. This causes a steady stream of discovery permission alerts.

Today sql MP is recommended for max 50 databases. That's not very enterprise-ish. Maybe a complete rewrite is needed.

It seems there is no easy way to determine why a SQL Cluster is not being discovered. The host OS gets discovered, but if you don't have all the accounts right and sometimes it still doesn't get discovered. there doesn't seem to be anything in the logs to say why its not being discovered.