npm run build: Build the source codenpm run dev: Watch for changes and recompile files on changes
Before running profiles or benchmarks, make sure to run npm run build first to compile the source code to JavaScript, compiled code will be put in dist folder.
Benchmarking scripts are located in folder src/benchmarks
node dist/benchmarks/benchmark_connect_multiple_endpoints.js -l [library] -n [numberOfNetworks]
Where:
[library]is the library that will be used to connect:polkadotapi,dedotordelightfuldot-poc[numberOfNetworks]is the number of network endpoints to connect (5,10,20,50,100)
E.g 1: Connect to 100 of network endpoints via delightfuldot
node dist/benchmarks/benchmark_connect_multiple_endpoints.js -l delightfuldot -n 100
E.g 2: Connect to 100 of network endpoints via polkadotapi
node dist/benchmarks/benchmark_connect_multiple_endpoints.js -l polkadotapi -n 100
The benchmark was running on a Macbook Pro M1 / 32GB RAM - 512GB Storage, NodeJS v18.14.2
The memory consumption can be varied over time depending on various reasons. Below numbers are the average results of 5 consecutive runnings.
| Number of network endpoints | @polkadot/api | delightfuldot |
|---|---|---|
| 5 | ~ 50 MB | ~ 18 MB |
| 10 | ~ 105 MB | ~ 25 MB |
| 20 | ~ 210 MB | ~ 40 MB |
| 50 | ~ 470 MB | ~ 90 MB |
| 100 | ~ 830 MB | ~ 140 MB |
Profiling scripts are located in folder src/profiles
Connect to Polkadot using @polkadot/api
node --inspecting dist/profiles/profile_connect_to_polkadot_via_polkadotapi.js
Connect to Polkadot using delighfuldot
node --inspecting dist/profiles/profile_connect_to_polkadot_via_delightfuldot_poc.js