About Machinebeat

This Beat is experimental and may be changed or removed completely in a future release. Do not use in production.

The current version in master of Machinebeat is a well working and tested prototype. It will be developed further over time but can already be used. The downloadable binaries are expected to run without any mayor issues. The current state is experimental. Breaking changes are possible in future releases!

What is a beat? A beat is a lightweight data shipper written in GOLANG developed by Elastic N.V. and the community. It is open source and also implements a framework offered to the community to build their own beats. Those beats may offer special purpose data collection not offered by existing beats. Machinebeat is one of those special purpose beats.

The aim of machinebeat is, to build a leightweight data shipper that is able to pull metrics from machines used in industrial environments and ship it to Elasticsearch or have Logstash or a message queue like Kafka inbetween, depending on the use case and architecture.

Today Machinebeat is able to connect to OPC-UA Servers, MQTT Broker and IoT Clout Services like AWS IoT core. It can pull data in real-time, subscribing to topics and feeding the data into Elasticsearch for near real-time monitoring purpose. This will enable machine operators to see behavior of their machines and analyse important metric on the fly. If you want to collect data from PLCs check this out: PLC4X at Github

The ability to get machine metrics and other related information is a foundation for maintenance, quality assurance and optimization of industry 4.0 environments. It can help to see possible issues early in the production cylce or optimize the use of incredients for the products produced. There are many other problems that can be solved monitoring machine metrics once they can be derived from them.

In a future version Machinebeat is supposed to support additional protocols in order to cover a broader mix of different sensor metrics in the industry.

Latest release news 23. October 2020

The OPC UA module browsing is adding the path for the node now. This helps to indentify nodes during ingest for enabling ingest pipeline features. Ingest pipelines helping e.g. to modify or drop data that is not necessary.

This version also supports running multiple modules within one instance of Machinebeat.

The OPC UA module now creates only 1 subscription per instance. This one subscription handles all nodes that you need to get data from.

Let's test it

You don't have an Elastic cluster up and running? Use Elasticsearch Service

Getting Started with Machinebeat

To start quickly and easy use the pre build binaries shared in this repo! To get started choose the version that fits for your environment.

ARM

Download the latest version of the binary here: https://ela.st/machinebeat-arm

Linux

Download the latest version of the binary here: https://ela.st/machinebeat-linux

Start the beat with the following command in your terminal.

./machinebeat -e -c machinebeat.yml

Use -e to see the log output. Use -c to set a different config file.

Windows

Download the latest version of the binary here: https://ela.st/machinebeat-windows

Start the beat with the following command in your CMD or PowerShell. There is also a PowerShell Script to add the beat as a service.

machinebeat.exe -e -c machinebeat.yml

Use -e to see the log output. Use -c to set a different config file.

Configurations

OPCUA Module

Machinebeat uses the GOPCUA project to connect and pull data or subscribe from OPS-UA servers. GOPCUA is an implementation of the OPC-UA specification written in GOLANG. More about OPC-UA can be found on Wikipedia

To enable the OPCUA Module rename the file modules.d/opcua.yml.disabled to modules.d/opcua.yml

The current version is able to read and subscribe nodes from an OPC-UA address space specified in the modules.d/opcua.yml file from Servers and transfer the collected data to Elasticsearch directly or via Logstash. The nodes and its leafs will be discovered automatically mentioned YAML file like this:


- module: opcua
  metricsets: ["nodevalue"]
  enabled: true
  period: 2s
  
#The URL of your OPC UA Server
  endpoint: "opc.tcp://opcuaserver.com:48010"

Different Nodes can be specified and read by machinebeat.

MQTT Module

To enable the MQTT Module rename the file modules.d/mqtt.yml.disabled to modules.d/mqtt.yml Change the configuration based on your needs. It works with every MQTT Broker and also IoT Cloud Services.

Example configuration to collect from AWS IoT core:

- module: mqtt
  metricsets: ["topic"]
  enabled: true
  period: 1s
  host: "tcps://<yourAWSEndpoint>:8883/mqtt"
  clientID: "<yourAWSclientID>"
  topics: ["#"]
  decode_payload: true
  ssl: true
  CA: "<pathToAWSRootCA>"
  clientCert: "<pathToAWSyourIoTCertificate>"
  clientKey: ""<pathToAWSyourIoTPrivateKey>""

Your client id from IoT console -> things:

arn:aws:iot:us-east-2:<AWS account>:thing/<clientID>

Make sure your certificate has correct policies attached:

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Action": "iot:*",
      "Resource": "*"
    }
  ]
}

How to build on your own env

1.) Download all dependencies from go.mod using go get -u

2.) You may need to overwrite some modules with the following versions that do not support go.mod in older versions

go get k8s.io/client-go@kubernetes-1.14.8
go get k8s.io/api@kubernetes-1.14.8
go get k8s.io/apimachinery@kubernetes-1.14.8

3.) Run go build in the machinebeat repository

barisvelioglu / Machinebeat