We are going to implement parallel-safe AVL trees using coarse-grained locks, fine-grained locks and lock-free mechanism using C++. We will analyze the cost and speedup of each implementation on a variety of workloads to draw conclusions on the advantages and disadvantages of each approach.
| Week | Date | Task | In-charge |
|---|---|---|---|
| 1 | 3/25 - 3/31 | Create correctness/speed tests and workload sets | Matt |
| 2 | 4/1 - 4/7 | Finish coarse-grained lock implementation | Matt |
| 3 | 4/8 - 4/14 | Finish fine-grained lock implementation | Tra |
| 4 | 4/15 - 4/21 | Enforce correctness, optimize performance, and write milestone report | Both |
| 5 | 4/22 - 4/28 | Finish lock-free implementation | Both |
| 6 | 4/29 - 5/5 | Conduct performance analysis and write final report | Both |