Skip to main content

MongoDB Metrics

This tutorial shows you how you can visualise MongoDB metrics in SigNoz. We take an example of a FastAPI sample app which uses MongoDB to walk through this tutorial.

Based on how you are using MongoDB, your exact steps may be different - but this should give you an idea of how to go about it.

Install FastAPI sample app via docker​

  1. Clone this sample FastAPI application

Follow instructions in the README file to set up the application. Please ensure to point the FastAPI application to IP of the machine where SigNoz is installed

  1. Use 172.17.0.1 - if running SigNoz also in same VM as the FastAPI application
http://172.17.0.1:4317

This comes with Mongo Exporter built in which exposes metrics in prometheus format at port 9216

Check that MongoDB metrics are exposed at following end point​

curl http://localhost:9216/metrics

Update Otel Collector config file to scrape MongoDb metrics​

Add a job name corresponding to mongodb exporter in otel-collector-metrics config file in your SigNoz install

signoz/otel-collector-metrics-config.yaml

- job_name: "mongo-collector"
          scrape_interval: 30s
          static_configs:
            - targets: ["172.17.0.1:9216"]

Job names should be aligned

Please ensure that the edited YAML is correctly configured. You can use tools like http://www.yamllint.com/ to check the correctness of the updated YAML file

Make changes in otel-collector-metrics file

Restart Otel Collector metrics container​

docker-compose --env-file ./docker/clickhouse-setup/env/x86_64.env -f docker/clickhouse-setup/docker-compose.yaml restart otel-collector-metrics

check that otel-collector-metrics is running by doing

sudo docker ps

Plotting Mongo metrics in SigNoz​

Good guide on metrics you may want to monitor

Create new panels in Dashboard section of SigNoz and add the following queries for example. 

PromQL query - mongodb_ss_connections
Legend Format - {{conn_type}}

PromQL query - rate(mongodb_ss_opcounters[5m])
Legend Format - {{legacy_op_type}}
tip

You can use queries from the json files here to plot more metrics

Grafana Dashboard for MongoDB exporter

https://grafana.com/grafana/dashboards/2583/revisions