Which architectural elements support reliable telemetry delivery across multiple sites?

• A. A single gateway device without redundancy.
• B. Edge processing, gateway device redundancy, and centralized message broker with failover.
• C. Only cloud-hosted solutions with no local processing.
• D. Direct public internet without security controls.

Answer: (B)

Reliable telemetry across multiple sites depends on resilience in data collection, processing, and transport. The best approach combines edge processing, gateway redundancy, and a centralized message broker with failover.

Edge processing brings data handling closer to the source, allowing local filtering, aggregation, and even initial decision-making. This keeps essential telemetry flowing even if the link to the central system is slow or intermittent, reduces bandwidth needs, and lowers latency for time-sensitive actions.

Gateway device redundancy ensures there isn’t a single point of failure at the data ingress. If one gateway goes offline, another can continue collecting and forwarding telemetry, maintaining continuity of data flow across sites.

A centralized message broker with failover provides durable, reliable transport between producers and consumers. It decouples data generation from processing, buffers messages during outages, and, with failover across brokers or clusters, keeps ingestion and delivery continuous even if part of the system experiences issues.

In contrast, relying on a single gateway without redundancy creates a fragile entry point that can halt telemetry if it fails. Depending solely on cloud-hosted processing with no local handling increases vulnerability to connectivity problems and latency. Bypassing security controls and sending data directly over public internet introduces risk and instability, undermining reliability and trustworthiness of the telemetry pipeline.