Which two headend router scalability factors should be considered when designing a DMVPN network solution that uses a hub-and-spoke topology? (Choose two.)

The aggregated packets per second rate is a direct headend scalability limit - the router must encapsulate and decapsulate traffic from all active spoke tunnels simultaneously, and exceeding the platform's PPS ceiling causes packet drops regardless of available bandwidth.

Bandwidth is a WAN circuit sizing concern rather than a headend router hardware scalability factor - it describes the uplink capacity, not the router's ability to process and maintain tunnel state.

The routing protocol affects convergence behavior and design complexity but is not a measurable hardware scalability constraint on the headend router platform itself.

The maximum number of tunnels is a hard platform limit; exceeding the router's NHRP/IPsec tunnel table capacity will prevent additional spokes from registering, making it a primary design constraint when estimating spoke growth.

CPU and memory are underlying resources that contribute to scalability but are not the actionable scalability metrics - PPS capacity and maximum tunnel count are the specific limits derived from those resources and used in design calculations.