meli

README

meli

Meli is supposed to be a faster alternative to docker-compose. Faster in the sense that, Meli will try to pull as many services(docker containers) as it can in parallel.

Meli is a Swahili word meaning ship; so think of Meli as a ship carrying your docker containers safely across the treacherous container seas.

It's currently work in progress.

I only intend to support docker-compose version 3+; https://docs.docker.com/compose/compose-file/compose-versioning/

Meli is NOT intended to replicate every feature of docker-compose, it is primarily intended to enable you to pull, build and run the services in your docker-compose file as fast as possible.
If you want to exec in to a running container, use docker; if you want to run an adhoc command in a container, use docker; if you want..... you get the drift.

Installing

Very early test releases are available from the releases page

Usage

meli --help

Usage of meli:
  -up
    	Builds, re/creates, starts, and attaches to containers for a service.
        -d option runs containers in the background
  -f string
    	path to docker-compose.yml file. (default "docker-compose.yml")
  -v	Show version information.
  -version
    	Show version information.

cat docker-compose.yml

version: '3'
services:
  redis:
    image: 'redis:3.0-alpine'
    environment:
      - RACK_ENV=development
      - type=database
    ports:
      - "6300:6379"
      - "6400:22"

meli -up

2017/10/07 14:30:11 {"status":"Pulling from library/redis","id":"3.0-alpine"}
2017/10/07 14:30:11 {"status":"Digest: sha256:350469b395eac82395f9e59d7b7b90f7d23fe0838965e56400739dec3afa60de"}
2017/10/07 14:30:11 {"status":"Status: Image is up to date for redis:3.0-alpine"}
...
2017/10/07 14:30:12 
u2017-10-07T11:30:12.720619075Z  1:M 07 Oct 11:30:12.720 * The server is now ready to accept connections on port 6379

Build

git clone git@github.com:komuW/meli.git
go build -o meli main.go
./meli -up -f /path/to/docker-compose-file.yml

Benchmarks

Aaah, everyones' favorite vanity metric yet no one seems to know how to conduct one in a consistent and scientific manner.
Take any results you see here with a large spoon of salt; They are unscientific and reek of all that is wrong with most developer benchmarks.

Having made that disclaimer,

test machine:
lsb_release -a

Distributor ID:	Ubuntu
Description:	Ubuntu 16.04 LTS
Release:	16.04
Codename:	xenial
No LSB modules are available.

uname -ra

4.4.0-96-generic #119-Ubuntu SMP Tue Sep 12 14:59:54 UTC 2017 x86_64 x86_64 x86_64 GNU/

docker-compose version:
docker-compose --version

docker-compose version 1.16.1, build 6d1ac219

Meli version:
version 0.0.5

Benchmark test:
this docker-compose file

Benchmark script:
for docker-compose:
docker ps -aq | xargs docker rm -f; docker system prune -af; /usr/bin/time -apv docker-compose up -d
for meli:
docker ps -aq | xargs docker rm -f; docker system prune -af; /usr/bin/time -apv meli -up -d

the above scripts were ran 3 times for each tool and an average taken.

Benchmark results(average):

tool	User time(seconds)	Elapsed wall clock time(seconds)
docker-compose	1.61 sec	63.57 sec
meli	0.04 sec	28.43 sec

Thus, meli appears to be 2.2 times faster than docker-compose(by wall clock time).
There are still some low hanging fruits, performance wise, that I'll target in future.
But I'm not making a tool to take docker-compose to the races.