Other monitoring and logging solutions

Cloud Service provides a Prometheus-compatible endpoint you can use to connect to your own monitoring infrastructure as well as Postgres logs by way of blob storage.

If you're using your own cloud account, contact Cloud Service Support to enable this feature on your clusters.

Metrics

You can access metrics in a Prometheus format if you request this feature from Cloud Service Support. You can retrieve the hostname and port for your clusters by using the Prometheus URL available on the Monitoring and logging tab on each cluster's detail page in the Console.

These example metrics can help you get started.

Patterns for accessing metrics

A common pattern for metric shipping is to have the vendor-supplied agent scrape the metrics endpoint and send the metrics to the desired platform.

For more information on some common monitoring services, see:

• Dynatrace Prometheus extension

• Self-managed Grafana (not Grafana Cloud): Grafana Prometheus datasource

Logs

You can view your logs in your cloud provider's blob storage solution if you request this feature from Cloud Service Support. You can retrieve the location of your object storage on the Monitoring and logging tab on your cluster's detail page in the Console.

The general pattern for getting logs from blob storage into the cloud provider's solution is to write a custom serverless function that watches the blob storage and uploads to the desired solution.

Watching for logs on AWS

You can leverage some Python code to read the S3 bucket and then use the AWS APIs to upload to your custom monitoring solution. For example, for CloudWatch: aws-load-balancer-logs-to-cloudwatch.

Watching for logs on Azure

You can leverage Azure functions to read the Azure Blob Storage object. Then use the API to upload the data to your monitoring solution.

Uploading logs to common third-party providers

After your function observes new log data, you can use your monitoring provider's API to push log data to their platform.

Some platform providers have limitations regarding the ingestion of logs. Read the vendor documentation carefully.

• Datadog

  • AWS: Collecting logs from S3 buckets

  • Azure: Log collection from Blob Storage

• Dynatrace

  • AWS: S3 log forwarder

  • Azure: Azure log forwarder

    Warning

    Currently, the Dynatrace integration works only if you can stream the logs from Azure Storage to Azure Event Hub.

• New Relic

  • AWS: Lambda for sending logs from S3

  • Azure: Send logs from Azure Blob storage

CLI command

You can also get the metrics and logs URLs using the cluster show-monitoring-urls CLI command. See Logging and metrics CLI command for more information.