Kubernetes

Introduction

Important

Not for production use!!

This guide explains how to deploy the Open AMT Cloud Toolkit services in Microsoft Azure. Scaling functionality in MPS enables Open AMT Cloud Toolkit to support a greater number of managed devices. For this deployment, kubernetes runs in Microsoft Azure along with redis and consul. Redis is used to sync the Web Server sessions between Web Server instances. Consul is used to communicate device connections between the Web Server and MPS server.

High level Design

Figure 1: MPS Scaling Architecture

Figure 1 presents the high-level architecture of MPS scaling mode. Per the figure, starting at the bottom:

1. Devices connect to an available MPS Server through the load balancer.
2. The REST API requests are routed to an available Web Server, a component of MPS running in scale mode, through a load balancer.
3. The Web Server determines which MPS Server to route the traffic to based on which MPS Server the device is connected to and sends that traffic through the MPS Proxy connection.
4. The MPS Server sends the traffic to the corresponding device.

MPS Configuration

To support running the service in a distributed environment, some configuration settings were added to MPS. These settings can be modified in open-amt-cloud-toolkit\scripts\kubernetes\serversChart\values.yaml. All the following settings have already been preset in the values.yaml file.

Table 1: MPS configuration settings

Component:	Setting:	Notes:
consul	distributed_kv_name	name of key/value store used.
	distributed_kv_ip	ip of key value store
	distributed_kv_port	port used by key/value store
redis	redis_enable	enable redis caching for web session
	redis_host	redis host
	redis_port	redis port
	redis_password	password used to authenticate to redis
general	web_proxy_port	port the web server used to communicate to MPS.
	network_adaptor	network identifier used when device connects to MPS. Can be either an adaptor name such as eth0 or starting ip address such as 192.168.
	startup_mode	microservice run mode. standalone when running in non scaling mode or run components in mps and web for distributated mode.

Get the Toolkit

To clone the repositories:

1. Open a Command Prompt or Terminal and navigate to a directory of your choice for development:

git clone --recursive https://github.com/open-amt-cloud-toolkit/open-amt-cloud-toolkit

Building the docker images

Please build and push the following images from the following open-amt-cloud-toolkit repositories MPS, RPS, and sample-web-ui into a public repository. Please update "docker.io/vprodemo" to the desired public repository and "mps" to the corresponding repository name.

docker build . -t docker.io/vprodemo/mps:latest
docker push docker.io/vprodemo/mps:latest

Prerequisite

Install required software in Windows environment(in admin mode):

1. Run \open-amt-cloud-toolkit\scripts\kubernetes\installchoco.bat to install the choco package manager
2. Close and reopen command window
3. Run \open-amt-cloud-toolkit\scripts\kubernetes\installpackages.bat to install the required packages

Steps to Deploy

1. Edit entries in the launch.bat file to fit deployment.

2. Update section images in stack environmental variables in open-amt-cloud-toolkit\scripts\kubernetes\serversChart\values.yaml that correspond with desired images for launch.

3. If containers are in a private container registry, a base64 auth token needs to be created and placed in open-amt-cloud-toolkit\scripts\kubernetes\config.json

4. Launch with command launch.bat [resourceGroupName] [replace the resourceGroupName with the desired resource group name for the deployment]

5. Use the settings in the \open-amt-cloud-toolkit\scripts\kubernetes\.env file to update the values in \open-amt-cloud-toolkit\scripts\kubernetes\serversChart\values.yaml

6. From \open-amt-cloud-toolkit\scripts\kubernetes run helm install openamtcloudstack ./serversChart to deploy the services into kubernetes.

7. After deployment is complete use the command kubectl get pods to verify all pods have been launched successfully.

Note

Please restore values.yaml to its prelaunch condition before subsiquesnt deployments