All notable changes to this project will be documented in this file.
- lock dependencies, upgrade urllib3 to fix security vulnerability
- fix filtering logic to support conditions where labels are not present
- Add Support for Sumo Logic Prometheus Content Type
- update pipfile.lock
- set X-Sumo-Client header
- Add ability to remove labels before sending to Sumo Logic
- Refactor callback to allow callback to be optional and invoke callback per metric as opposed to per batch.
- harden config validation
- add discovery capabilities for metrics behind Kubernetes services
- allow verify to be a boolean or string to allow skipping verification
- refactor
- remove print statement
- Add ability to include/exclude metrics based on labels.
- Fix bug that could cause filtering by metric name not to behave as expected.
- Better error handling to reduce volume of logging.
- Add ability to pass callback function to SumoPrometheusScraper
- major refactor, which introduces some breaking changes for the better, hence the major version bump
- use Pipfile for dependencies management and repeatable builds
- include README.md, CHANGELOG.md, LICENSE and other files in the built image
- wildcard support in include_metrics / exclude_metrics
- stricter config validation
- allow sumo_http_url to be defined in global config or overridden per target
- allow configuration to use environment variables
- clean up and refactoring
- switch to alpine image
- ensure job and instance are part of dimensions so they are included with all metrics
- build in scheduling logic and allow targets to specify the interval they should run.
- reject NaN values
- async calls for targets and sending to sumo, add retry logic for posting to sumo.
- Add support for verify cert and token.
- Add retry logic to all target calls.
- Better error handling on request to target for data.
- Add ability to include metrics explicitly, in addition to current exclusion logic in each target definition.
- If metric value is NaN, convert to 0.
- Add target name, used to generate up metric to show that service is up and healthy.
- Always send up metric.
- Initial Release