fig_17_unikraft-sqlite-libc

SQLite Performance Experiment

Goal

This folder is used for a performance experiment of SQLite running on Unikraft.

This experiment compares four different configuration:

• newlib-native uses the native port of SQLite to Unikraft, with the newlib libc and pthread-embedded.
• musl-native uses the native port of SQLite to Unikraft, with musl libc.
• musl-compat uses a static library of SQLite built with musl-gcc on a standard GNU\Linux box (we recommend Debian), which is then linked with Unikraft, also using musl.
• linux-native does not use Unikraft at all. It just runs the same benchmark running SQLite built with musl-gcc, as a standard Linux process (no virtualisation).

All Unikraft experiments use the TLSF memory allocator, while the Linux one uses the default Linux allocator.

How to Use

Running the sqlite_experiments.sh as a user will:

• create a temporary folder;
• download SQLite, Unikraft and the relevant Unikraft libraries;
• configure, build and run the different test applications (10 times each).

For the Linux test, if /tmp is on tmpfs, i.e. a RAM disk, it will be used. Otherwise the experiment will create a RAM disk. It is important to use a RAM disk to be fair as the Unikraft experiments are using the RAM disk approach, and the benchmark is I/O intensive. Note that on some Linux boxes, using /tmp on tmpfs is actually faster than using a dedicated RAM disk mounted using tmpfs as well (sync operations return immediately with /tmp while provoking context switches with other RAM disks).

After execution of the sqlite_experiments.sh script, the temporary folder will contain the built files. You can inspect the applications and re-run their experiments using the run.sh in the apps folders, or by calling sqlite_experiments.sh --run.

Note that with no information about the system, those run.sh scripts assume CPU3 to be available, and only fix the CPU frequency governor to performance (e.g. you are still responsible for isolating that CPU, disabling turboboost, etc.). It is up to you to make that setup if needed. Both the main sqlite_experiments.sh and individual run.sh scripts are easy to modify directly.