What are two advantages of using model-driven telemetry, such as gRPC, instead of traditional telemetry gathering methods? (Choose two.)

Model-driven telemetry is subscription-based and structured using YANG data models, not ad-hoc; ad-hoc queries are characteristic of traditional polling methods like SNMP GET requests.

gRPC uses Protocol Buffers for binary serialization, which is significantly more compact than text-based formats like XML used in NETCONF or SNMP MIB data, resulting in substantially lower bandwidth consumption per telemetry message.

gRPC and streaming telemetry do introduce protocol overhead including connection establishment, subscription negotiation, and encoding; the advantage is reduced overhead compared to polling, not zero overhead.

Model-driven telemetry does not inherently provide decentralized storage; telemetry data is streamed to a designated collector or pipeline and storage architecture is a separate design concern independent of the telemetry protocol.

Model-driven telemetry supports dial-in and dial-out subscription models that push incremental state changes to the collector as soon as they occur, providing near-real-time visibility without the latency and overhead of periodic polling cycles.