Scaling of SCCM hierarchy is very important with the release of new SCCM Current Branch with SCCM servicing and Windows 10 servicing options. Planning and designing of the infrastructure is very important with SCCM CB. We need to carefully and smartly identify when we might need to use additional distribution points or management points. Most of us, who already used SCCM CAS or read about the SCCM CAS related advice from the community will try to avoid the CAS server installation. Rather we will try to flatten the SCCM CB Infra with stand alone primary server. Now flattening the SCCM Infra is very much possible with the increase of supporting numbers of clients; a SCCM Current Branch stand alone primary site supports the following number of devices: 175,000 total clients and devices. SCCM Current Branch Scale out options should be very dynamic in an enterprise level, for that 3rd party SCCM add on like Adaptive Onesite will become very handy.
Standalone primary server can support 15 management points and 250 DPs. So with flatter infrastructure we may have to deploy loads of DPs and MPs in the environment. I’m not very big fan of secondary sites for low bandwidth sites mainly because SQL DB replication. More details in the following post here. But for large environments, we may have trouble in placing MPs/DPs in each remote offices mainly because SCCM CB can support only 15 MPs, 250 DPs and 250 secondary sites. We can’t even think of operation teams overhead if we have hundreds of DPs and secondary sites in your SCCM CB infra. TechNet documentation on SCCM current Branch supported sizing and scale numbers here. To scale out your SCCM Current branch architecture and design you must look into smart scaling options like Adaptiva Onesite. Adaptiva can dynamically handles remotes offices without spanning out hundreds of DPs across your WAN. More details about Adaptiva here.