MKIt - Micro Kubernetes on Ruby

This is micro kubernetes(tm) on Ruby(tm), a simple tool to deploy containers to mimic a (very) minimalistic k8 cluster with a nice REST API.

It contains a internal DNS and uses HAProxy for routing/balancing/fail-over for Pods access. The database is a simple sqlite3 db and the server is a Sinatra based application.

The client is not done yet, only includes a simple wrapper to curl

The daemon is responsible for HAProxy pods routing configuration. It also provides the cluster DNS and manages the internal host interface.

Requirements

• Ruby
• HAProxy
• Docker
• Linux (iproute2 package)

Install

This is a simple ruby gem, so to install run

# gem install mkit

Running

The daemon requires root user (due to ip and haproxy), you can run it directly on the repository root...

# ./mkitd  --help
Usage: mkitd [options]
    -c config-dir                    set the config dir (default is /etc/mkit)
    -p port                          set the port (default is 4567)
    -b bind                          specify bind address (e.g. /tmp/app.sock)
    -s server                        specify rack server/handler
    -q                               turn on quiet mode (default is off)
    -x                               turn on the mutex lock (default is off)
    -e env                           set the environment (default is development)
    -o addr                          set the host (default is (env == 'development' ? 'localhost' : '0.0.0.0'))

or after the gem install mkit-<version>.gem. The server and client will be installed on host.

# mkitd
...
 0.65s     info: MKIt is up and running! [ec=0xbe0] [pid=45804] [2023-12-29 15:46:04 +0000]

There's also samples on the samples dir, for daemontools and systemd.

Accessing the API

• Create new service
  • mkitc create samples/apps/rabbitmq.yml
• Update service
  • mkitc update samples/apps/rabbitmq.yml
• Get service
  • mkitc ps {id|service_name}
• Delete service
  • mkitc rm {id|service_name}
• List services
  • mkitc ps [-v (verbose)]
• Control service
  • mkitc start {id|service_name}
  • mkitc stop {id|service_name}

Example:

$ mkitc ps postgres
id      name                 addr             ports                      status
4       postgres             10.210.198.10    tcp/5432                   RUNNING
  pods
    id      pod_id            pod_name          pod_ip            status
    19      4ce31a007211      5d148a16f3aa      172.17.0.2        RUNNING

The service postgres is available on IP 10.210.198.10:5432

Configuration

On startup, configuration files on config directory will be copied to /etc/mkit. HAProxy config dir and control commands are defined on mkit_config.yml

You must configure haproxy to use config directory. e.g. on Ubuntu

# /etc/default/haproxy

# Defaults file for HAProxy
#
# This is sourced by both, the initscript and the systemd unit file, so do not
# treat it as a shell script fragment.

# Change the config file location if needed
CONFIG="/etc/haproxy/haproxy.d"

# Add extra flags here, see haproxy(1) for a few options
#EXTRAOPTS="-de -m 16"

Service

service:
  name: rabbitmq # unique
  image: rabbitmq:3-management-alpine # image
  network: bridge # docker network - it will be created if it does not exists
  ports:  # haproxy port mapping: <external_port>|<internal_port>|<tcp|http>|round_robin
    - 5672:5672:tcp:round_robin
    - 80:15672:http:round_robin
  resources:
    max_replicas: 1
    min_replicas: 1
  volumes:
    - docker://mkit_rabbitmq_data:/var/lib/rabbitmq # a docker volume - it will be created if it does not exists
    - /var/log/rabbitmq/logs:/var/log/rabbitmq # a local volume
  environment:
    RABBITMQ_DEFAULT_USER: admin
    RABBITMQ_DEFAULT_PASS: admin
    RABBITMQ_DEFAULT_VHOST: mkit

Development

• build the gem
  • rake package
• console
  • rake console

Thanks

For my kids. :)