Your research orchestrator dispatches five worker agents in parallel to retrieve data from different…

Question

Your research orchestrator dispatches five worker agents in parallel to retrieve data from different sources. Three workers return successfully within 8 seconds, but two are still pending at the 30-second mark with no response. The orchestrator must decide whether to wait or proceed. Which strategy best balances result completeness with system reliability?

Accepted Answer

B. Set a per-worker timeout; if a worker exceeds it, mark that source as unavailable in the report and proceed with results from completed workers, flagging the incomplete coverage explicitly.

Answer

A. Always wait for all workers to complete before proceeding, regardless of elapsed time, to ensure the report has complete source coverage.

Answer

C. Cancel all pending workers and restart the entire parallel dispatch with a fresh set of worker agents.

Answer

D. Promote one of the slow workers to a "blocking" dependency and apply timeouts only to the remaining pending worker.

Question 239