The document discusses how Blackboard sizes its Academic Suite software based on benchmarking. It provides details on the benchmarking methodology, including modeling user behavior, data growth, and performance objectives. The results showed how the software performed under different workload levels on various hardware configurations. The last part discusses using the benchmark results and sizing guide to determine an institution's adoption profile and appropriate hardware configuration based on factors like sessions per hour and page loads.
1. How We Size the Academic Suite: Benchmarking at Blackboard TM Speaker: Steve Feldman Director, Software Performance Engineering and Architecture [email_address]
7. The Performance Lifecycle Complete End to End Performance Engineering Refactoring and Optimizing End to End Performance Testing Modeling, Profiling and Simulation Data Collection & Usage Analysis Strategy, Methodology and Best Practices SPE Methodology
17. Performance Scenarios Identical workload to the under-loaded Learning System with Community System model, but with the definition of 50 complex domain relationships. Response times calibrated to under-loaded system comparison (~5s.) Calibrated Academic Suite with Complex Domains Combination of Learning System and Community System use case interactions with 40% of the workload in a controlled Assessment Concurrency Problem. Response times calibrated to under-loaded system comparison (~5s.) Calibrated Learning System and Community System with Concurrency Model for Assessments Combination of Learning System, Community System and Content System use case interactions to reflect the budding adoption of the full Academic Suite. Response times calibrated to under-loaded system comparison (~5s.) Calibrated Academic Suite Regression test case from 6.3 performing a mix of student viewing/activity, instructor authoring and minimal administrator management. Meant to be an over-loaded system. Response times < 15s. Over-Loaded Learning System and Community System Regression test case from 6.3 performing a mix of student viewing/activity, instructor authoring and minimal administrator management. Meant to be an calibrated loaded system. Response times < 10s. Calibrated Learning System and Community System Regression test case from 6.3 performing a mix of student viewing/activity, instructor authoring and minimal administrator management. Meant to be an under-loaded system. Response times < 5s. Under-Loaded Learning System and Community System Summary/Description Workload
18.
19. Performance Objective #1 58% 2650 minutes (4 threads) 6360 minutes Large Institution (Sun Microsystems) 58% 130 minutes (4 threads) 309 minutes Moderate Institution (Sun Microsystems) 36% 16 minutes (4 threads) 25 minutes Small Institution (Sun Microsystems) Improvement Benchmark #2 (Min. Threads) Benchmark #1 Model Name NA Not Valid 989 minutes Large Institution (Windows) 40% 2120 minutes 5389 minutes Large Institution (Linux) NA Not Valid 196 minutes Small Institution (Windows) 37% 107 minutes (4 threads) 288 minutes Moderate Institution (Linux) NA Not Valid 9 minutes Small Institution (Windows) 12 minutes (4 threads) 21 minutes Small Institution (Linux) Improvement Benchmark #2 (Min. Threads) Benchmark #1 Model Name
42. Storage 3:1 Greater then 1 TB 600,000 500,000 4:1 800 GB 300,000 50,000 5:1 200 GB 50,000 5,000 10:1 20 GB 7,000 500 Ratio of File System to Database Storage File System Size Number of Existing Users Number of Existing Courses