MiCADO - autoscaling framework for Docker services on Cloud

This software was developed as part of the now completed COLA project. The University of Westminster (U.K.) and MTA SZTAKI (Hungary) will continue its open-source development as part of other research projects.

MiCADO is an auto-scaling framework for Docker applications. It supports autoscaling at two levels. At virtual machine (VM) level, a built-in Kubernetes cluster is dynamically extended or reduced by adding/removing Nodes hosted on cloud virtual machines. At container level, the number of replicas implementing a Docker Service is automatically increased/decreased. The application detailing the services, links and scaling rules must be specified by a TOSCA-based Application Description Template (ADT).

The MiCADO manual is hosted at https://micado-scale.readthedocs.io .

Manuals for MiCADO versions are as follows, and there is a very basic quick start guide below:

• MiCADO v0.9.2
• MiCADO v0.9.1-rev1
• MiCADO v0.9.1
• MiCADO v0.9.0
• MiCADO v0.8.0
• MiCADO v0.7.3
• MiCADO v0.7.2-rev1
• MiCADO v0.7.2
• MiCADO v0.7.1
• MiCADO v0.7.0
• MiCADO v0.6.1
• MiCADO v0.6.0
• MiCADO v0.5.0

Quick Start Guide

Requirements

1x MiCADO Master

• 2 CPU / 4GB RAM / 20GB DISK
• Ubuntu 18.04 or 20.04
• On a supported cloud
  • AWS EC2
  • CloudSigma
  • CloudBroker
  • OpenStack
  • Microsoft Azure
  • Google Cloud
  • Oracle Cloud Infrastructure

1x Ansible machine (could be your laptop, or a VM instance)

• Ansible >= v2.9
• curl
• jq (for demos)
• wrk (for demos)

Set-up

Clone the repository & fill in connection details and credentials. In the hosts file, complete values for ansible_host (Master IP) and ansible_user (SSH user). In the credentials- files provide usernames & passwords according to cloud, desired MiCADO WebUI access and/or private container registry:

MICADO_PATH=/home/ubuntu/micado
git clone https://github.com/micado-scale/ansible-micado $MICADO_PATH
PLAYBOOK_PATH=$MICADO_PATH/getmicado/playbook

cd $PLAYBOOK_PATH
git checkout v0.9.2

cd $PLAYBOOK_PATH/inventory
cp sample-hosts.yml hosts.yml
edit hosts.yml

cd $PLAYBOOK_PATH/project/credentials
cp sample-credentials-cloud-api.yml credentials-cloud-api.yml
cp sample-credentials-micado.yml credentials-micado.yml
edit credentials-cloud-api.yml
edit credentials-micado.yml

#optional for private registry login#
cp sample-credentials-docker-registry.yml credentials-docker-registry.yml
edit credentials-docker-registry.yml

Run the micado.yml playbook to completion:

cd $PLAYBOOK_PATH
ansible-playbook project/micado.yml

Now view the MiCADO Dashboard at https://<MiCADO_Master_IP>

Testing

Prepare a demo by filling _settings with your MiCADO Master IP and user/pass:

cd $MICADO_PATH/demos/stressng
edit _settings

Fill the required fields (ADD_YOUR_ID ...) under the worker-node virtual machine with IDs from your cloud service provider, then run the test scripts:

cp stressng_<yourcloud>.yaml stressng.yaml
edit stressng.yaml
./1-submit-tosca-stressng.sh
./2-list-apps.sh
./3-stress-cpu-stressng.sh 85

Check out the Dashboard to see the test scale up, and undeploy when done:

./4-undeploy-stressng.sh

Edges

Using the sample-hosts-with-edges file as a reference, fill values under the micado key as before, pointing at the MiCADO master. Under the edgenodes key, add as many uniquely named edge devices as desired, specifying their IPs as ansible_host and SSH username as ansible_user

cd $PLAYBOOK_PATH/inventory
cp sample-hosts-with-edges.yml hosts.yml
edit hosts.yml

Run the edge.yml playbook to completion:

cd $PLAYBOOK_PATH
ansible-playbook project/edge.yml

When viewing the MiCADO Dashboard, check Nodes under the Kubernetes Dashboard to confirm your Edge device(s) are connected