Puma key performance indicators - Application Monitoring with Amazon CloudWatch

Puma key performance indicators

This solution deploys a dashboard for Puma, which allows you to monitor the following key performance indicators by default:

  • Puma traffic patterns.

    • Puma Traffic by RemoteIP: Top 10 Remote IPs

    • Puma Traffic by Host: Top 10 Hosts

  • Puma throughput and latency metrics to report on the health of individual servers and detect bottlenecks or other performance problems.

    • Puma Traffic by ResponseTime: The response time for requests

    • Puma Traffic by BytesSent: Bytes sent in response to requests

    • Puma Traffic by BytesReceived: Bytes received in request payloads

  • Puma Host-level resource metrics to determine whether to scale servers up or out.

    • Host mem_used_percent: The percentage of memory currently in use

    • Host tcp_established: The number of TCP connections established

    • Puma CPU: CPU utilized by bundle process

  • Puma status to identify misconfigured files or other underlying issues.

    • Puma Traffic by Status: HTTP status for requests

  • Puma server metrics to identify application performance on server.

    • Pool Capacity: The number of maximum concurrent connections possible

    • Max Threads Count: The number of maximum threads allowed at any time

    • Running Threads Count: The number of threads currently running

    • Workers Count: The number of workers currently running

    • Requests Count: The number of requests made to Puma web server

  • Custom Puma metrics supported by StatsD to measure application performance.

    • Page Load Time: Time elapsed to load page

    • Page Hit Count: The number of times page accessed