Goal: - [ ] Performance Overhead due to remus alone (no libvmi) Apps/workloads: - [ ] Httperf - Varying workloads, varying checkpoints Metrics: - [ ] Throughput and latency of Httperf Graphs: - [ ] checkpoint vs throughput (constant workload) - [ ] checkpoint vs latency (constant workload) - [ ] Select an ideal checkpoint that gives max throughput and a reasonable latency - [ ] workload vs throughput (constant/ideal checkpoint) - [ ] workload vs latency (constant/ideal checkpoint) Key points:
Goal:
Apps/workloads:
Metrics:
Graphs:
Key points: