What logging practice is essential for traceability in the EPD Pilot?

• A. Unstructured logs with no identifiers
• B. Timestamped, structured logs that include message IDs, sequence numbers, endpoints, outcomes, and error codes for traceability
• C. Only log errors without timestamps
• D. Store logs locally without central collection

Answer: (B)

Traceability in logging comes from capturing each action with enough context to follow a transaction across systems. The best practice is to use timestamped, structured logs that include a message ID, sequence number, the endpoint involved, the outcome, and any error codes. This combination lets you link related events from different services, order them correctly, see exactly where a step occurred, and quickly understand whether it succeeded or failed and why.

Timestamps provide the timeline so you can reconstruct the sequence of events. Structured fields make it easy to search, filter, and correlate data across logs and services. A unique message ID ties all log entries to the same transaction, while a sequence number preserves the order of events within that transaction. Endpoints show where in the system something happened, outcomes reveal success or failure, and error codes give precise clues for troubleshooting. Centralized log collection then enables cross-system traceability, audits, and long-term analysis.

Unstructured logs with no identifiers make it hard to connect events across services. Logging only errors, without timestamps, deprives you of timing and context. Storing logs locally without central collection prevents you from aggregating and analyzing logs from multiple components, breaking traceability.