CMSC624 Database Architecture and implementation final project. Haoran Zhou, Tianrui Guan, Junran Yang
https://dl.acm.org/doi/pdf/10.1145/3318464.3389764
https://github.com/zhr1201/STRIFE-public/blob/master/CMSC624_final_project.pdf
We implemented two modified STRIFE schedulers which work better than the original STRIFE under certain workloads. But we haven't got the time to reproduce the throughput under the original workloads in the paper.
We deleted those parts because they are parts of the assigment of the CMSC624 course https://github.com/abadid/624Assignment2. If you want to finish those parts, please also don't make this public!
The serial clusterer can work fine but there are known memory corruption issues with the parallel clusterer. We didn't fix it cause we are not working on the project at least for now.
This makefile orgnization doesn't support multi-folder C++ projects very well. Including header files from other directory in the src files and test files can lead to redifinition or missing implmentation issues. Probably should switch to https://github.com/zhr1201/ParrallelSudokuSolver (a Kaldi style makefile) for continuing working on this project.
We decided not to work on the project for now.
Branching: create ur own branch from develop, merge into and merge from the develop branch frequently to prevent diverging too much from the develop branch.
Adding source file: when adding a source file, we need to add the xxx.cc file in the Makefile.inc in the corresponding subdir.
Adding tests: It's better to keep your unit tests in the xxx_test.cc file. Please don't write private tests and then just drop them once they are passed. In order to run the tests, we need to first have a xxx.cc file and add it to SUB_DIR_NAME_SRC and then name the test file as xxx_test.cc.