Redis Queue/List Length Metrics Exporter

Pull code:

git clone https://github.com/ashishtiwari1993/redis_queue_length_exporter.git
cd redis_queue_length_exporter

Edit queue.json:

{  
  "queue_type":{  
	"queue_group1":[  
	  "q1",
	  "q2",
	  . 
	  .
	],
	"queue_group2":[  
	  "q1",
	  "q2",
	  . 
 	  .
	],
	"queue_group3":[  
	  "q1",
	  "q2",
	  . 
 	  .
	],
	"queue_group4":[  
	  "q1",
	  "q2",
	  . 
 	  .
	]
  },
  "hosts":{  
	"host1.example.com:6379":[  
	  "queue_group1",
	  "queue_group2"
	],
	"host2.example.com:6379":[  
	  "queue_group3",
	  "queue_group4"
	]
  }
}

Define queue group in queue_type object. Queue group has array of all queue name (q1, q2) of same behaviour. Map Queue group with host in host object. It will export queue length accroding to queue.json file and It will assign queue labels in prometheus while storing data.

Here is the sample file queue.json.

Run locally:

pip install prometheus_client
pip install redis
python llen.py

Run via Docker:

docker build -t "redis_queue_length_exporter" .
docker run  -d --name 'redis_queue_length_exporter' -p 8000:8000 redis_queue_length_exporter

Test on browser:

localhost:8000/metrics

Add Job to the Prometheus config

Add below block to the prometheus.yml file.

 - job_name: redis_llen_exporter
   scrape_interval: 5s
   scrape_timeout:  5s
   metrics_path: "/metrics" 
   static_configs: 
   - targets: ['localhost:8000']

Prometheus will pulling the data for every 5 seconds from the localhost:8000/metrics.

You can change port, interval and timeout according to your requirements.

Grafana Dashboard

ID: 10287
Dashboard URL: https://grafana.com/grafana/dashboards/10287

What's exported?

Queue length from the LLEN command are exported, See https://redis.io/commands/llen for more details.

What else?

• Open an issue if you have more suggestions or ideas about what to add.