unity2promgo

Prometheus Exporter for DellEMC Unity

Unity2Prom is a tool written in Go for exporting usage and performance metrics from Dell EMC Unity storage arrays to prometheus.

For RestAPI connectivity to the different Unity arrays gounity by Equelin (fork by muravsky) is used. Modifications related to Docker incorporated from cthiel42 Modifications related to compile go code at docker build phase from pedrocrc

This version support LUN polling to be able to match LUN names to ID's

Configuration

Configure the Exporter

Before starting the exporter it is required to update config.json with values specific to your environment:

{
  "exporter": {
    "port":     8080,
    "interval": 60,
    "pools": true,
    "storage_resources": true,
    "LUN": true,
    "metrics": ["sp_cifs_global_basic_writeBytesRate", "sp_cifs_smb1_usage_currentConnections","sp_net_device_pktsInRate","sp_net_device_pktsOut"]
  },
  "unitys": [
    {
      "ip": "192.168.2.190",
      "port": "",
      "user": "monitor",
      "password": "Monitor123!"
    }
  ]
}

Unitys* is a list of all arrays for which data will be collected. For the user it is sufficient to be a Operator User with read-only access.

Exporter conifgures the exporter and specifies the data that will be collected. !!!The interval in which metrics are collected should match the Prometheus scraping interval!!! Metrics can be added by adding the metric's prom_path to the metrics array. Possible metrics and their description can be found in unity_metrics.json:

  {
    "id": 10170,
    "name": "Read_Hit_Only Miss requests",
    "path": "sp.*.blockCache.flu.*.readHitOnlyMiss",
    "prom_path": "sp_blockCache_flu_readHitOnlyMiss",
    "type": 3,
    "description": "Number of Read_Hit_Only Miss requests",
    "isHistoricalAvailable": false,
    "isRealtimeAvailable": true,
    "unitDisplayString": "requests"
  },

Configure Prometheus Scrape Job

In order to configure Prometheus to scrape the newly added metric it is requried to add another scrape_config entriy to the promehteus configuration file:

scrape_configs:
- job_name: unity2go  
  scrape_interval: 60s
  scrape_timeout: 10s
  metrics_path: /metrics
  scheme: http
  static_configs:
  - targets:
    - <unity.fqdn|ip_adress>:9090    

Installation

Compile the collector In the project base directory run go build main.go utils.go unitycollector.go

Start the collector In the project base directory run ./main

Docker

Ensure that the configuration file you have is up-to-date and accurate with the metrics you want collected, the destination of your Unity array, and the credentials to connect to the array. Then you will be able to build the image.

Build the image with the following command (alpine)

docker build . --file Dockerfile.x86_64.alpine --tag unity2promgo

Build the image with the following command (debian)

docker build . --file Dockerfile.x86_64.debian --tag unity2promgo

Run the container with the following command

docker run -d \
        --name unity2promgo \
        --restart=always \
        --net=host \
        unity2promgo

An alternative way to run the container would be to pull the image from dockerhub with the following command

docker pull pedrocrc/unity2promgo:alpine

and then mount your configuration file to the container in your docker run command similar to below

docker run -d \
        --name unity2promgo \
        --restart=always 
        --net=host \ 
        -v /Path/to/config.json:/opt/unityexporter/config.json:ro \ 
        pedrocrc/unity2promgo:alpine