Benchmarks Writer CLI (#6567)

* initial mockup

* add and wipe

* track writes

* start to add to pipeline

* return all reads/writes

* Log reads and writes from bench db

* causes panic

* Allow multiple commits

* commit before ending benchmark

* doesn't work???

* fix

* Update lib.rs

* switch to struct for `BenchmarkResults`

* add to output

* fix test

* line width

* @kianenigma review

* Add Whitelist to DB Tracking in Benchmarks Pipeline (#6405)

* hardcoded whitelist

* Add whitelist to pipeline

* Remove whitelist pipeline from CLI, add to runtime

* clean-up unused db initialized whitelist

* Add regression analysis to DB Tracking (#6475)

* Add selector

* add tests

* debug formatter for easy formula

* initial idea

* use all benchmarks

* broken

* working without trait

* Make work for multiple pallets

* Fix merge issues

* writer appends to file

* implement () for balances weight trait

* update name of trait

* Weights to WeightInfo

* auto trait writer

* Heap pages are configurable

* clean out runtime changes

* more clean up

* Fix string generation

* Update comments

* Update bin/node/runtime/src/lib.rs

Co-authored-by: arkpar <arkady.paronyan@gmail.com>
This commit is contained in:
Shawn Tabrizi
2020-07-06 11:34:24 +02:00
committed by GitHub
parent a9c21b8b84
commit 2019f70768
8 changed files with 266 additions and 36 deletions
+5 -5
View File
@@ -22,11 +22,11 @@ use linregress::{FormulaRegressionBuilder, RegressionDataBuilder, RegressionMode
use crate::BenchmarkResults;
pub struct Analysis {
base: u128,
slopes: Vec<u128>,
names: Vec<String>,
value_dists: Option<Vec<(Vec<u32>, u128, u128)>>,
model: Option<RegressionModel>,
pub base: u128,
pub slopes: Vec<u128>,
pub names: Vec<String>,
pub value_dists: Option<Vec<(Vec<u32>, u128, u128)>>,
pub model: Option<RegressionModel>,
}
pub enum BenchmarkSelector {