Propagate downstream disconnects through streamed responses
A cancelled report download continues consuming upstream bandwidth.
- Focused work estimate
- 3h + prerequisites
- Priority in the scenario
- High
- Engineering practice
- Streams · Cancellation
Estimated field mix
- Networking50%
- Backend30%
- Performance engineering20%
Field percentages are editorial estimates of the ticket's engineering focus. They total 100%; they are not measured time, proficiency scores, or ownership evidence.
Review it, then add it to your workspace.
The board opens an editable draft; nothing is saved until you confirm it. Sign-in and workspace permissions apply, and Demo boards remain ephemeral.
Project context
A fictional reporting API sits behind a reverse proxy. Clients see intermittent disconnects and incorrect origins after changes to keep-alive and forwarding headers.
Setup prerequisites
- Create two loopback HTTP services and a local proxy with synthetic requests; no public scanning or external traffic.
Preceding work
Complete these dependencies, or supply their agreed outputs before taking this ticket.
- BPROXY-101 · Document the trusted proxy chain and public origin
- BPROXY-102 · Normalize forwarded headers without accepting client spoofing
- BPROXY-103 · Strip hop-by-hop headers before forwarding requests
- BPROXY-104 · Align keep-alive lifetimes between proxy and upstream
- BPROXY-105 · Limit upstream connections per service without unbounded waiting
Acceptance criteria
- Abort the owned upstream request on disconnect.
- Stop buffering and release stream resources.
- Preserve completion metrics distinct from cancellation.
Implementation constraints
- Use bounded synthetic report streams.
Verification to include
- Finish a small streamed response.
- Disconnect mid-stream and verify upstream cancellation.
Deliverables
- Streaming cancellation fix.
Rollout and recovery
Adopt on one streaming route; retain traces of safe lifecycle events.
Value of the work
For the engineer: Practice HTTP intermediaries, connection lifetimes, and network failure isolation.
For the team: Provide a proxy contract that preserves request identity and releases resources predictably.
Evidence boundaries
Outcome Evidence: Tests, patches, and runbooks are requested deliverables. They become Outcome Evidence only through a qualified Mission and immutable Evidence IDs.
Ownership Evidence: Independent adaptation must be observed under a declared verification policy and cite immutable Evidence IDs. Completing a planning ticket establishes no Ownership Evidence.