Skip to content

docs(outputs): document output_drop deployment-stream event#227

Draft
maximelb wants to merge 2 commits into
masterfrom
docs/output-drop-deployment-event
Draft

docs(outputs): document output_drop deployment-stream event#227
maximelb wants to merge 2 commits into
masterfrom
docs/output-drop-deployment-event

Conversation

@maximelb
Copy link
Copy Markdown
Contributor

Summary

  • Adds documentation for the new output_drop deployment-stream event emitted by the endpoint when an output drops records (saturation or retries exhausted).
  • Describes the event's routing/payload shape, fields, and behavioral guarantees (aggregation window, self-suppression on the deployment stream and on tmp_live_ outputs, final-flush on drain).
  • Updates the deployment-event types table and "Use Cases" list to surface output-health monitoring as a use case.

Pairs with the corresponding endpoint change in legion_endpoint-go (feat/credit-backpressure) that introduces the retry helper and drop aggregator.

Test plan

  • mkdocs build renders the new section cleanly
  • Field table and JSON example match what emitDropEvent in endpoint/storage/output_retry.go actually publishes
  • No broken cross-links

🤖 Generated with Claude Code

maximelb and others added 2 commits May 13, 2026 13:06
When a user-configured output cannot keep up or its retries are exhausted,
the platform now emits an aggregated output_drop event on the deployment
stream so operators can detect and alert on data loss without parsing
internal logs.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Implementation detail, not user-facing.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant