Download presentation
Presentation is loading. Please wait.
Published byRenata Domagała Modified over 6 years ago
1
Simulating a $2M Commercial Server on a $2K PC
A.R. Alameldeen, M.M.K. Martin, C.J. Mauer, K.E. Moore, M. Xu, D.J. Sorin, M.D. Hill, D.A. Wood Presented By: Derek Hower
2
Contributions Develop a cost and time efficient simulation methodology for multiprocessor systems. Tuned and scaled benchmarks Dealing with variability Extended timing simulator
3
Workload Tweaking Wisconsin Commercial Workload Suite
OLTP – On-Line Transaction Processing SPECjbb – Java Middleware Apache – Static Web Server Slashcode – Dynamic Web Server Scaled to reduce memory and disk usage Tuned on an actual multiprocessor server to discover bottlenecks
4
Case Study: OLTP Based on TPC-C v3.0, using IBM DB2 V7.2 EEE
Scaled to 3 sales districts per warehouse, 30 customers per district, and 100 items per warehouse Compared to 10, 30,000 and 100,000 required by TPC Set up on a Sun E5000 Disk images were moved to simulator
5
Case Study: OLTP cont Initial Scaling - Kernel/device tuning
Reduced entire simulation to fit in 1GB of memory ( MB warehouses) Kernel/device tuning Changed limits on semaphore usage, threads, locks, etc Database separated from kernel and spread out over 5 physical disks Reducing contention increased # of warehouses, keeping db size constant
6
Case Study: OLTP cont Additional Concurrency Added more users
7
Simulation Shorten simulations as much as possible while still maintaining accuracy Start with warm workloads using snapshots Fixed simulation length based on # of transactions Account for variability by introducing random memory access delays and by averaging multiple simulation runs
8
Timing Added proc and memory timing models to Simics Memory model:
Timing-first simulation Memory model: cache coherence cache latencies and bandwidth memory interconnection network
9
Evaluation Simulated system using Bandwidth Adaptive Snooping Hybrid (BASH)
10
Thoughts Validation Mentioned briefly but skirted the issue Can we trust the data? Is there a loss of generality when scaling and tuning workloads?
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.