Open tareknaser opened 7 months ago
Reorganize benchmarks into groups to mirror crates. For now, I think we should have benchmarks for two crates:
data-resource
fs-index
Currently, our benchmarks suite look like this:
../test-assets/lena.jpg compute_bytes ../test-assets/test.pdf compute_bytes compute_bytes_large compute_bytes compute_bytes_medium compute_bytes compute_bytes_small compute_bytes index_build
We should have benchmarks structure similar to this:
index_query get_resource_by_id get_resource_by_path track_addition track_removal update_index resource_id_creation compute_from_bytes:large compute_from_bytes:medium compute_from_bytes:small compute_from_path:../test-assets/lena.jpg compute_from_path:../test-assets/test.pdf
We should add a note for "Generating Profiles for Benchmarks" in our README To resolve issue https://github.com/ARK-Builders/ark-rust/issues/40 We should add flamegraph as the suggested method.
README
Description
Reorganize benchmarks into groups to mirror crates. For now, I think we should have benchmarks for two crates:
data-resource: These benchmarks evaluate the performance of resource ID creation from file paths and random data (already implemented).fs-index: These benchmarks assess the performance of resource indexing queries and operations.Implementation Details
Currently, our benchmarks suite look like this:
We should have benchmarks structure similar to this:
Profiling
We should add a note for "Generating Profiles for Benchmarks" in our
READMETo resolve issue https://github.com/ARK-Builders/ark-rust/issues/40 We should add flamegraph as the suggested method.