Skip to content

[AWS API Gateway] Metrics datastream #6270

@lucian-ioan

Description

@lucian-ioan

Summary

Collect metrics, analyze them and produce an integration for API Gateway metrics. Analyzing these metrics helps detect bottlenecks, optimize resources, and troubleshoot issues.

This integration will include all three supported API types: HTTP, REST, and WebSocket.

Tasks

  • Set up or use existing instances for the three API types to generate metrics
  • Initially inspect metrics for confirmation via AWS CloudWatch or AWS CLI
  • Fetch metrics via metricbeat and save sample event
  • Create the integration in accordance to fetched data which includes fields, dimensions, sample event, icon etc.
  • Spin up the stack locally; Set up and verify the integration via Kibana
  • Add unit and metric types
  • Create documentation which includes summary, setup, requirements, fields and sample event

Challenges

  • AWS naming scheme can be confusing (ex: 4XXError for REST vs 4xx for HTTP). Both metrics exist and appear to serve the exact same purpose but it's easy to assume one is just a typo initially.
  • Given that no metrics are currently being generated in the Elastic account for API Gateway, metricbeat might not fetch metrics consistently. This is because it misses the time window that the CloudWatch metricset uses.

Docs

Metrics and dimensions documentation:

Unit and metrics types documentation:
https://www.elastic.co/guide/en/elasticsearch/reference/master/mapping-field-meta.html

Metadata

Metadata

Assignees

Type

No type
No fields configured for issues without a type.

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions