This folder will accumulate many examples of operations with its benchmark. With time, it will be added several languages together with its Mongodb library.
The Mongodb used is local installation whether system/current os installation or using Docker to run it in order to avoid network cost timing. We are measuring the lib itself after all, not our network speeds.
For using Docker, we can simply run
docker run --rm -dp 27017:27017 mongo:{version}with version is the Mongodb version, such as latest or 4.2, or others e.g.
docker run --rm -dp 27017:27017 mongo:4.2
The Python is using virtual environment (venv) to avoid polluting workspace.
To setup the venv, do
python3 -m venv .
./script/activate
pip install pymongo
pip install motor # used in insertfind examplesto create a Python environment in this particular folder and installing pymongo library.
For Nim examples, it's depended on benchy package so install with nimble
nimble install benchyand then we can simply run it like
nim r -d:danger testfile.nimwe can add various compile options such as --gc:arc or --gc:orc to see its
difference.
As long we have the Golang installation ready, to test we do
go mod tidy
go test -benchmem -bench .and look for third tab, that has somenum ns/op, as the timing.
For this particular (benchmark) examples, any contribution is appreciated to give
a nuanced and fair comparison for each language version.
By no means the values are absolute as each app/program for each user's case is
different,
but this should give better information and perspective.
$ nim -v
Nim Compiler Version 1.4.4 [Windows: amd64]
Compiled at 2021-02-23
Copyright (c) 2006-2020 by Andreas Rumpf
$ python -V
Python 3.9.6
$ go version
go version go1.16.4 windows/amd64$ nim r -d:danger --gc:orc ./{examplename}.nim
$ python ./{examplename}.py
$ go test -bench . -benchmem| Test\Language | Nim (ms) | Python (ms) | Golang (ms) |
|---|---|---|---|
| insert bulk/many | 92.040 | 151.54565 | 118.215250 |
| insert 100 (1 by 1) and find last | 78.683 | 41.523589 | 80.01250 |