You have to provision 400 devices and you need to distribute them among three IoT hubs, using Device Provisioning Service. All the hubs are linked to the DPS and you need to design how the devices are distributed over the different hubs. Here are the requirements for the different types of assets: Simple temperature and CO-sensors: can be attached to any of the 3 hubs, with no preference
Edge devices: their data must be queried from a registry database; allocation on the basis of the result
Surveillance cameras: the images must be available real time in case any actions need to be taken
Which allocation policy would you use for the given scenarios?

Answer options:

A.Static; Custom; Lowest latency
B.Default; Custom; Lowest latency
C.Evenly weighted; Lowest latency; Custom
D.Evenly weighted; Custom; Static

Correct Answer: B
Option A is incorrect because for the simple sensors the default allocation policy, i.e. the evenly weighted distribution is the perfect choice.
Option B is CORRECT because The default allocation policy (i.e. Evenly weighted) is the right selection for the simple sensors for which no specific rule applies; edge devices need a custom logic to be executed before deciding the allocation hub, therefore the “Custom” policy built as Azure Function App is suitable; telemetry data from cameras must reach the IoT hub with no delay, so the “Lowest latency” policy is the best choice.
Option C is incorrect because the requirement for the edge devices can be fulfilled by the custom policy (built as a Function App); the cameras must be allocated focusing on the latest latency. (The simple sensors can be distributed evenly over the hubs, which is the default setting, so it is correct.)
Option D is incorrect because for the cameras requiring real-time visibility, the lowest latency setting must be used.
References: 

https://docs.microsoft.com/en-us/azure/iot-dps/tutorial-custom-allocation-policies?tabs=powershell
https://docs.microsoft.com/en-us/azure/iot-dps/how-to-use-custom-allocation-policies?tabs=windows