Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
test(tm2/pkg/cmap): add benchmarks to show true impact of contention (g…
…nolang#3540) The Go standard library's sync.Map is touted as great for cases with high load and is commonly known knowledge but the benchmark that I am committing shows otherwise that for this library's usage, it is so much more expensive hence this benchmark will avoid someone committing sync.Map without seeing the true implications. ```shell $ benchstat map_w_mutex.txt stdlib_sync_map.txt name old time/op new time/op delta CMapConcurrentInsertsDeletesHas-8 1.72s ±11% 1.92s ± 3% +11.66% (p=0.000 n=10+9) CMapHas-8 109ns ± 9% 118ns ± 3% +8.26% (p=0.002 n=10+8) name old alloc/op new alloc/op delta CMapConcurrentInsertsDeletesHas-8 1.18GB ± 2% 3.21GB ± 3% +172.09% (p=0.000 n=10+10) CMapHas-8 16.0B ± 0% 16.0B ± 0% ~ (all equal) name old allocs/op new allocs/op delta CMapConcurrentInsertsDeletesHas-8 824k ± 0% 4433k ± 0% +437.89% (p=0.000 n=10+10) CMapHas-8 2.00 ± 0% 1.60 ±38% ~ (p=0.065 n=9+10) ``` Updates gnolang#3505
- Loading branch information