I would like to have additional logging when Distribution points get updated. Currently the update process is not really verbose if you look it from the SCCM Console point of view.
I had an issue with dll's not being upgraded in the DP Upgrade process and there was no trace of the error.

Thanks

When going through the process of changing the configuration for the site database (in order to move it to a new SQL server), if you are keeping your SQL server name the same, the wizard will not allow you to input the same SQL server name as the old. We want to get rid of our SQL 2008 R2 servers running on Windows 2008 R2 and replace with Server 2016 and SQL 2016, but also keep the same hostname for the new server. Our only option since the wizard doesn't support this scenario is to do a backup and restore.

MS Case : 17052664 >> Post upgrade of SCCM CB 1906 we realised that one of remote SMSPROVIDER server was not working properly. This remote SMSPOVIDER never had ADK installed on it since SCCM infrastructure was setup. Error never reported in the past in any upgrade but it stopped working post successful upgrade of 1906. In order to fix this issue in future it should be part of Prerequisite check in upgrade setup.

Currently the Database contains the build number and shows the base version such as 5.00.8498.1000 where the updated version is 5.00.8498.1700 or 1702 (currently in my case).

It cant be hard to store the "Friendly" version information in the database.

You could even break it down into Site version, Console Version and Client version if necessary if these components are likely to get out of step due to hotfixes.

Service accounts should be domain aware so that they do not attempt to authenticate against the wrong domain to lock out service accounts with the same username but on another domain.

With the ability to support 250 DPs and 2000 PDPs on a site, it would be helpful to be able to see the site systems on a map. Currently, the geographical view only accommodates site servers. DPs also make good sense, since they are typically geographically dispersed and would make a pretty map to show off to management.

When Group Discovery for the whole Domain is activated and there are Organizational Units containing the Slash Character in the Name, you get "failed to bind to Container" Errors in SCCM.

As long a Slash is an allowed character in LDAP, SCCM should be able to process these OUs correctly.

For those running a Tech Preview lab, you may forget sometimes to update your lap and your release may expire.

In production sites, it would also be nice to trigger the upgrades on a schedule to not have to be in front of the console to start the upgrade.

It would be great if you could schedule the servicing upgrades to be ran on a schedule based on release date. Start X days after release at X time, much like an ADR. You could also expose this to PowerShell to provide more customization.

Perform permission checks on key locations where sccm will be setup. We had a problem experienced a few years ago where we had a "hash mismatch" error b/c the account I was using to perform the upgrade did not have permission to the install directory.

https://sccmf12twice.com/2017/02/sccm-2012-r2-sp1-failed-upgrade/

this problem is still possible in CB

Ability to support more clients than 150k on a single Primary site. Our environments could get rid of bottlenecks and less administration overhead if there is no need to deploy a CAS. It would be great if a Primary site supports up to 200k clients, because we have currently 160k in our environment.

When clients upload Information to a Secondary Site MP, that Information gets consolidated, compressed and then sent up in the hierarchy by ONE single file, in order to minimize WAN issues.
However at standalone MPs each small Client communication is being sent up using one little file, resulting in many little packages consuming lots of unnecessary WAN bandwidth.
That prevents us from getting rid of our secondary sites even though we urge to do so, i.e. because they caused critical failure when updating to ConfigMgr 1511 (classified by Microsoft as known Bug)

Some enterprises are disabling the use of SHA-1 certificates in their environment. To avoid using a SHA-1 certificate for the Data Replication Service (DRS), you must move to SQL 2017 where the self-signed certificate defaults to SHA-256. The use of a PKI certificate would provide more flexibility/security going forward.

At the moment if you double click a Status Message Query, it edits it, So to run it you need to right click it or select it and click the button on the ribbon.

It would be better if this was reversed so that double clicks run the query and to edit it you right click it

The Components which were installed during the SCCM Setup should be updated during the Current Branch process if it's an necessary feature for the Environment.

For Example the SQL Native Client. The most Companies got an separated SQL Team and got an Problem of responsibility during the Upgrade process.

Would like the ability to configure location of OfflineImageServicing location.