-
Task
-
Resolution: Done
-
Critical
-
None
-
None
-
None
-
False
-
False
-
NEW
-
NEW
-
Undefined
-
-
Logging (Core) - Sprint 201, Logging (Core) - Sprint 202
Task
Review & update the internal data model documentation and organization.
Goals
- Treat the data model as a contract or protocol between us and our users, and between our own components: collector, forwarder, normalizer, and store.
- Remove unused fields
- User Documentation: Generate documentation of the expected format of log records.
- Establish a process for the doc team to directly update the source
- Internal Documentation: Superset of user documentation, includes internal/private fields
- Classify fields as static or dynamic
- Static values are the same for every record in a log stream (for example Container-ID never changes over a container's life)
They can sometimes be encoded as part of a stream-id or attributes to avoid sending on every record: Loki labels, CloudWatch stream IDs, etc. - Dynamic values may be different in every record, so they must be sent each time.
- Static values are the same for every record in a log stream (for example Container-ID never changes over a container's life)
- Multi-encoding: The model can be encoded in multiple ways, add notes on mapping to JSON, Loki labels, Syslog STRUCTURED-DATA, etc.
Acceptance Criteria
- User documentation can be generated automatically from the model.
- A documented process for the doc team to update the model source.
- Removed out-of-date entries
- Classified remaining entries: static/dynamic, public/private
- is related to
-
RHDEVDOCS-2521 Document our JSON log entry format.
- Closed