Skip to content
/ yoda Public

Simple tool to dockerize and manage deployment of your project

Notifications You must be signed in to change notification settings

Muvon/yoda

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Yoda

Simple tool to dockerize and manage deployment of your project

Alt text

What is it?

Yoda helps you to dockerize existing application and automate deployment process.

  1. Only BASH. No dependency shit!
  2. Requirements: git, docker with compose plugin
  3. Its simple like simplicity itself
  4. Runs on MACOS and Linux systems

Installation

First you need to install Yoda on your laptop. Its supereasy:

git clone [email protected]:muvon/yoda.git
cd yoda && make check && sudo make install

Remember that you need bash version 4 or higher installed at least and GNU implementation of the sed.

Knowledge requirements

  1. Docker and its main concept
  2. Image and container understanding
  3. You know Docker Compose and its file syntax.

Usage example

OK. You have git repository with your project.
Go into this folder and run this command to initialize environment.

yoda init

Now you will get docker folder created in your project.
Next step is prepare Dockerfile that located in docker/images folder.
You can setup docker build options in file docker/Buildfile.

Now you can add container to your project.

yoda add container-name

Change template for docker-compose.yml file in docker/containers/container-name/container.yml.

We are done. Build it and start with just one command now:

yoda start

Done!

Philosophy

  1. You can have several images for single project.
  2. Each image you use must have Dockerfile located in docker/images folder and named by convention: Dockerfile-{name}.
  3. You can have several containers depends on one image.
  4. Each container has own folder in docker/containers with structure followed by convention in this README.
  5. You can setup and use any BASH variables in file docker/env.sh. Its pregenerated for you.
  6. Envfile is main file that has all info about what should be built and in which environment, also what server runs which environment for deploy.
  7. Each container can be scaled N times and started using single template but different names.
  8. You can fully customize deploy, build, compose stages just wrapping in your own script using any language.

Init folder structure

When you do yoda init in your project it creates by default yoda folder. This folder has following structure

Path Description
containers This folder contains all containers with templates in your project
images It contains Dockerfiles for you images. Common naming is: Dockerfile-name. Where is name is just name of image you build with that dockerfile
env.sh Its environment for your building context. You can define custom variables and use it everywhere in builds and other scripts. Its common for every environments you have. Optionally you can create special file for current environment. For example env.dev.sh. And this file will be read right after common file env.sh was read.
Buildfile It is declarative file that describes how to build each image you have. It has simple structure name: arguments for docker build command where is name is your image in images folder with same name. You can uses custom context also just in the format name[context]: arguments for docker build command. By default we use current dir as context for building
Envfile It describes all environments and link servers for deploy with its environments you have. No limitation. You can create own environments and describe what containers must be built there

Path: containers

When you adding new container folder is created here with same name. For example if you add container with name container. Same folder will appear here.
This folder will contain some files.

File Description
container.yml Its docker compose section without section name that describes how to build container. This file used to generate whole docker-compose.yml file for starting services
container.[env].yml Optional you can create file with template that will be used when docker-compose.yml is generated for this environment. For example if you have container.dev.yml and starting services in dev environment will use all keys from this file replacing common container.yml keys
entrypoint Its entrypoint for you container. Its optional but good practise to use this file as executable for your container starting point

What is container.yml

This file contains valid docker compose section for current service. container_name is immutable and declared by Yoda internally. You can specify image key here with shortcut to image from Buildfile. For example if your Buildfile describe image with key "base" you can put here just image: base and Yoda automatic will replace base to image from build params specified in Buildfile.

Path: env.sh

Here you can declare BASH environment variables and use it everywhere.

For example you can write here IMAGE_NAME to set image name with revision and other staff and use it in Buildfile and container.yml.

If you need custom env.sh file for environment you can just create it with name as env.dev.sh that will extend default env.sh file with new variables defined there.

Take a noticy that environment support namespaces for example production.server1 and production.server2. One server cannot hold 2 namespaces so to use custom env.sh in this case just use env.production.sh without namespace of each server.

Path: Buildfile

Its simple file that have following structure:

base: -t $IMAGE_BASE --compress
db: -t postgres:9.5

Each line contains image name and build args that will be passed in docker build command. See more info in docker build --help.

Path: Envfile

Its simple file YAML like with environment and server description:

user@server: production
production: container1 container2=2
dev: container1

Example file above declare server user@server that will be deployed as production. And production will contain one container1 and two container2 instances.
And in dev environment only one container with name container1 will be started.

Or in case of multiservers we use like this

user@server1: production.stack1
user@server2: production.stack2
production.stack1: container1 container2=2
production.stack2: container3 container4
dev: container1

In that case almost the same. Dot (.) allows us to separate namespaces and customize which services we use for one server and which for another. In that case we still have production environment. All that goes after dot (.) is about namespace of environment.

Path: Startfile

This file allow you to manage flow of start your complex service.

dev:
  flow: deploy container2 container1=2
  stop: container2
  wait: deploy
production.ns1:
  flow: deploy container2 container1=2
  stop: container2
  wait: deploy
production.ns2:
  flow: deploy container2 container1=2
  stop: container2
  wait: deploy
  • flow - It declares flow of starting services. It contains container or container=chunk where is chunk - number that divide starting by chunks
  • stop - each service represented here will be stopped before start
  • wait - After run up command with container sometimes you need to wait for exit code of it. This section declares which containers we should wait for

In this example you yoda start command first will stop container2 if its running.
After that it will start deploy, then container2 and then container1 services will be started as chunks by 2.
Also we wait for exit code for deploy container.

Take a notice in case if you use multiserver and namespaces in environment you have to describe each environment with namespace sepately because all of them independent in flow.

Path: docker/.yodarc

This is locked environment file for yoda inited in current project with yoda version and other useful common files. Please dont edit this file because it rewrites on yoda upgrade.

If you want you can redefine all variables in env.sh file.

Yoda command line tool usage

yoda command arguments

Commands available:

Command Description
version Display version of Yoda
help Display this information
init Prepare deployment folder in project
upgrade Upgrade to new version of initialized Yoda in project
add Add new container skeleton structure to project
delete Delete existing container from project
build Build images for current project
compose Display generated compose file for current environment
start Start all services for current project
stop Stop all services for current project
status Display current status of services
log Show log for a given container
logs Show log for all containers
exec Execute a given command in a container
enter Enter into a container with one of autodetected shell: zsh, bash or sh
setup Setup server to be read for use with yoda deployment system
deploy Deploy project on one or all nodes
rollback Rollback project on one or all nodes to previous revision
destroy Remove all created services by start command and all local images with volumes

yoda version

Display current Yoda version

yoda help

Display help information

yoda init [folder]

Prepare dockerized skeleton in project directory

Command Description Default
folder Initialize all structure in folder with that name yoda

yoda upgrade

Upgrade to new version of initialized Yoda in project.

yoda add [CONTAINER...]

Add container or bunch of containers skeleton to project

yoda delete [CONTAINER...]

Delete container or bunch of existing containers from project

yoda build [options] [IMAGES...]

Build images for current project. You can pass optional images you want to build. Default is every image from Buildfile. Options are:

Options Description Default
--rebuild Force build also if image exists already omitted
--no-cache Dont use Dockerfile cache when building images. omitted
--push Should we push built images to repository if we have REPOSITORY_URL defined in env.sh file omitted

yoda compose [COMPOSE_SCRIPT]

Display generated docker compose file in stdout.

Command Description Default
COMPOSE_SCRIPT Executable script who will process each container template, replace something and return as plain text. Container templates goes to stdin and 2 addition arguments are passed: --name and --sequence so name of container and number in scale map -

yoda start [options] [CONTAINER...]

Start all containers or only passed with arguments Options are:

Options Description Default
--rebuild Rebuild all images also if they exist with that revision omitted
--no-cache Dont use Dockerfile cache on building images stage. It passes internally to build command omitted
--recreate Force recreate containers
--force Should start containers excluding Startfile flow description

You also can manager flow of start and restart of you containers using Startfile

yoda stop [CONTAINER...]

Stop all containers or only passed with arguments

yoda status

Display current status of services

yoda log CONTAINER...

Show log for a given container

yoda logs

Show log for all containers

yoda exec CONTAINER... command

Execute a given command in a container

yoda enter CONTAINER...

Enter into a container with one of autodetected shell: zsh, bash or sh

yoda setup [options]

Setup server before it can be used for environment. You can use only Centos 8 distribution. Otherwise you have to setup it in manual mode. Before deploy put your authorized_keys to server folder.

Options Description Default
--env Setup single host or use host regexp pattern (Envfile will be used) -
--host Setup all servers in environment (Envfile will be used) -

yoda deploy [options]

Deploy single-node or whole cluster It exit with code 0 for success and 1 for failure (if something went wrong also on one node of all). Options are:

Options Description Default
--host Deploy on single host or using host regexp pattern (Envfile will be used) -
--env Deploy on all nodes with that environment (Envfile will be used) -
--stack Deploy only this stack in current environment -
--rev Set custom revision to be deployed or rollback to -
--branch What branch will be deployed. master
--args Custom environment arguments that will be passed to 'yoda start' command on each remote server to be deployed -
--force Pass this flag to yoda start command

yoda rollback [options]

Rollback to previous revision It exit with code 0 for success and 1 for failure (if something went wrong also on one node of all). Options are the same as for deploy command.

yoda destroy

Remove all created services by start command and all local images with volumes