benchmarking feed

feed_benchmarking.ex

Benchee.run(%{
  "Preload 1 query sync" => fn -> one_query |> Repo.all(in_parallel: false) end,
  "Preload 1 query async" => fn -> one_query |> Repo.all(in_parallel: true) end,
  "Preload 5 query async" => fn -> five_query |> Repo.all(in_parallel: true) end,
  "Preload 5 query sync" => fn -> five_query |> Repo.all(in_parallel: false) end
}, time: 10, memory_time: 2, warmup: 5)


Name                            ips        average  deviation         median         99th %
Preload 5 query async        191.25        5.23 ms    ±25.81%        4.96 ms       10.02 ms
Preload 1 query sync         184.57        5.42 ms    ±27.04%        5.21 ms        8.71 ms
Preload 1 query async        181.26        5.52 ms    ±32.14%        5.24 ms        9.21 ms
Preload 5 query sync         111.59        8.96 ms    ±69.42%        8.51 ms       16.53 ms

Comparison:
Preload 5 query async        191.25
Preload 1 query sync         184.57 - 1.04x slower
Preload 1 query async        181.26 - 1.06x slower
Preload 5 query sync         111.59 - 1.71x slower

Memory usage statistics:

Name                          average  deviation         median         99th %
Preload 5 query async       144.58 KB     ±0.02%      144.59 KB      144.59 KB
Preload 1 query sync        839.27 KB     ±0.05%      839.48 KB      839.62 KB
Preload 1 query async       839.33 KB     ±0.05%      839.48 KB      840.18 KB
Preload 5 query sync        344.37 KB     ±0.00%      344.37 KB      344.37 KB

Comparison:
Preload 5 query async       144.59 KB
Preload 1 query sync        839.48 KB - 5.81x memory usage
Preload 1 query async       839.48 KB - 5.81x memory usage
Preload 5 query sync        344.37 KB - 2.38x memory usage