• Overview
• Repository layout
• Docker configuration
• Setup scripts
  • MySQL Exporter
  • Linux Node Exporter
  • ActiveMQ Exporter
  • Windows / IIS / .NET Exporter
  • Windows / SQL Server Exporter

The goal of this project is to build a monitoring stack with Prometheus and Grafana inside Docker containers.

Both Prometheus and Grafana have multiple options and can be configured in many ways. In this repository we are using them to collect basic metrics from Microsoft Windows, IIS Web Services and SQL Server. As well as Linux operating system, ActiveMQ and MySQL services.

Prometheus is going to find which servers (targets) to monitor by using a service discovery integration. Discovery is available for AWS, Azure or Kubernetes, among others. Although in this repository we are using simple file discovery.

Metrics are exposed on target systems by exporters, providing an HTTP service which Prometheus can query at regular intervals. If a Docker daemon is available on the target system, exporters can run as containers as well. Otherwise, they can be installed manually as normal applications of course.

Along with Prometheus and Grafana we are also setting up Alertmanager for automated notifications.

This repository has four directories which contain all the required files.

• config: configuration files for supporting services
• dashboards: sample Grafana dashboards which work with our Prometheus configuration
• docker: required files to run Prometheus / Grafana in Docker containers
• scripts: scripts for setting up exporters on Windows / Linux

We are using docker-compose to run Prometheus / Grafana containers. This means all Docker configuration can be found in the docker-compose.yml file. Persistent data is stored on Docker volumes while configuration files are used through Docker bind mounts. To spin up a new monitoring stack you can run (from inside docker directory):

$ docker-compose up

Prometheus configuration and service discovery files can be found under the prometheus directory. Each file with the json extension is used by the file discovery service. The rules.yml file is used to setup alert conditions. And the prometheus.yml file is the main Prometheus configuration file.

Grafana configuration can also be found under the grafana directory. You will want to update the grafana.ini file. This is the main configuration file. Other files in this directory are used for automatic datasource provisioning.

The third service docker-compose will launch is Alertmanager. You can find its configuration under the alertmanager directory.

There's two required steps to use the mysqld_exporter. First, create a new database user which the exporter will use to connect to the MySQL server. And next, install the mysqld_exporter on the target system.

To create the database user, run the following commands using the mysql client.

CREATE USER 'mysqld_exporter'@'localhost' IDENTIFIED BY 'MySecret' WITH MAX_USER_CONNECTIONS 3;
GRANT PROCESS, REPLICATION CLIENT, SELECT ON *.* TO 'mysqld_exporter'@'localhost';

To install the mysqld_exporter, run the install-mysqld-exporter.sh Bash script.

To install the node_exporter, run the install-node-exporter.sh Bash script.

Apache ActiveMQ metrics are collected by Telegraf, a metrics collection agent with input/output plugins for many systems.

To install the Telegraf agent for ActiveMQ, run the install-telegraf-activemq.sh Bash script.

Microsoft Windows, IIS and .NET Framework metrics are collected by Telegraf, a metrics collection agent with input/output plugins for many systems.

To install the Telegraf agent for Windows / IIS / .NET, run the install-telegraf-dotnet.ps1 PowerShell script.

Microsoft Windows and SQL Server metrics are collected by Telegraf, a metrics collection agent with input/output plugins for many systems.

There's two required steps to use Telegraf with SQL Server. First, create a new database user which the exporter will use to connect to the SQL Server system. And next, install the Telegraf agent on the target system.

To create the database user, run the following T-SQL commands.

USE master;
GO
CREATE LOGIN [telegraf] WITH PASSWORD = N'MySecret';
GO
GRANT VIEW SERVER STATE TO [telegraf];
GO
GRANT VIEW ANY DEFINITION TO [telegraf];
GO

To install the Telegraf agent for Windows / SQL Server, run the install-telegraf-mssql.ps1 PowerShell script.