mirror of
https://github.com/GreptimeTeam/greptimedb.git
synced 2025-12-27 16:32:54 +00:00
d33435fa84
* feat: introduce wal benchmarker * chore: add log store metrics * chore: add some comments to wal benchmarker * fix: ci * chore: add more metrics for kafka logstore * chore: add more timers for kafka logstore * chore: add more configs * chore: move humantime to common dependencies * refactor: refactor wal benchmarker * fix: apply suggestions from code review * doc: add a simple README for wal benchmarker * fix: Cargo.toml * fix: clippy * chore: rename wal.rs to wal_bench.rs * fix: compile
Benchmarkers for GreptimeDB
Wal Benchmarker
The wal benchmarker serves to evaluate the performance of GreptimeDB's Write-Ahead Log (WAL) component. It meticulously assesses the read/write performance of the WAL under diverse workloads generated by the benchmarker.
How to use
To compile the benchmarker, navigate to the greptimedb/benchmarks directory and execute cargo build --release. Subsequently, you'll find the compiled target located at greptimedb/target/release/wal_bench.
The ./wal_bench -h command reveals numerous arguments that the target accepts. Among these, a notable one is the cfg-file argument. By utilizing a configuration file in the TOML format, you can bypass the need to repeatedly specify cumbersome arguments.