Updated 4/10/2018: Published a new version of the whitepaper (v1.1) - Corrected 4,000 to be 4,000,000 in section on SQL Express and Secondary Sites, additional pointers for WSUS related information added, other minor clarifications and updates.
We are sharing a draft, unofficial whitepaper with supplemental performance guidance for Configuration Manager environments of all sizes. We already document the maximum supported scalability limits as well as the hardware guidelines needed to run Configuration Manager (current branch) sites at the very largest environment sizes. This document includes specific details and test results generated by the product team, as well as frequently asked questions for common performance issues.
The largest contributor of performance bottlenecks in customer environments is the disk subsystem, so we provide suggested disk IOPS requirements for various size environments. Additionally, this paper documents a repeatable method that you can use to reproduce these numbers in your own environment and on your own hardware. This allows for more accurate estimates of the hardware required for Configuration Manager, whether on physical hardware, virtualized with Hyper-V on Windows Server, or running in Microsoft Azure.
Now we need your help! After review, please comment in the gallery discussion area, for example:
Do you have additional performance and scale questions?
Are there other best practices in your experience?
The whitepaper already includes information on Configuration Manager workloads such as inventory, discovery, collections. What are the top two other workloads that we should should prioritize for inclusion?
Were you able to successfully calculate disk IOPS for your environment?
Please share your feedback using the "Q and A" tab. We will incorporate your feedback when we finalize and publish this content in our official product documentation.