Agents schedule evidence gathering efficiently to meet deadlines
Before Agents Act: Assurance-Aware Semantic Scheduling for Evidence Acquisition in Distributed Systems
Distributed, Parallel, and Cluster ComputingArtificial IntelligenceCryptography and Security
Summary
Sometimes software agents need to collect proof before they make big changes to important systems, but the proof can become outdated if collecting it takes too long. The authors describe a new way of planning which evidence to get and when, so the information stays fresh and meets rules about diversity and deadlines. Their method greatly reduces stale evidence and makes plans that can adjust if problems arise. They tested their approach in simulated environments and showed it outperforms traditional scheduling methods.
What this means in practice
- •For infrastructure operators: Create efficient plans to gather valid and timely evidence before making system changes under resource and timing constraints.
- •For cloud service schedulers: Improve scheduling to ensure monitoring data remains fresh and diverse to support automated admission decisions within deadlines.
Tested on simulated data.
Authors
Jun He, Deying Yu
Abstract
Tool-using agents can initiate consequential infrastructure changes, yet evidence required for admission may expire while other checks run or depend on a shared fault domain. We formulate evidence acquisition as joint witness selection and scheduling under quorum, diversity, freshness, deadline, and resource constraints. Assurance-Aware Semantic Scheduling (AAS) combines integer-program selection, dispatch-aware temporal scheduling, bounded diagnostic expansion, and receipt-aware repair. Formal results state the assumptions needed for dispatch-time freshness and finite diagnostic expansion. In three generated infrastructure workloads, AAS produces 1,075/1,200 valid candidates versus 647/1,200 for constraint-aware forward scheduling; stale candidates fall from 440 to 12. Paired sensitivity studies reuse the same instances and operation latency draws across parameter settings. A corrected timeout intervention finds 18/20 admissions with repair or full resynthesis versus 0/20 for a static plan, with lower committed cost when receipts are reused. On 20 constructed cases requiring a certified decomposition cut, refinement recovers an oracle-matching feasible plan every time. These are controlled simulation results; the bounded oracle shares a temporal search component, and transfer to deployed systems remains untested.