peeramid-labs/quorum-rs
Watch
1
0
Fork
You've already forked quorum-rs
 0
Code Issues Pull requests Projects Releases 2 Packages Wiki Activity Actions

fix(providers): model health reads the catalog only, never a billed completion #95

Merged
peersky merged 1 commit from fix/catalog-only-model-health into dev 2026-09-01 07:37:44 +00:00
Conversation 0 Commits 1 Files changed 3 +29 -135
peersky commented 2026-09-01 07:37:31 +00:00
Owner
Copy link
  • Removed Interface ModelAvailability::with_serving_probe and probe_serving, which asked every seat's endpoint for one token on each probe interval
  • Fixed Bug where a fleet paid for a completion per seat every five minutes for the life of the process
  • Added Redundancy asserting the probe issues no request beyond the catalog fetch
  • Fixed Docs describing the proactive check as catalog-only, leaving a listed-but-unserved model to the reactive detector
- Removed Interface `ModelAvailability::with_serving_probe` and `probe_serving`, which asked every seat's endpoint for one token on each probe interval - Fixed Bug where a fleet paid for a completion per seat every five minutes for the life of the process - Added Redundancy asserting the probe issues no request beyond the catalog fetch - Fixed Docs describing the proactive check as catalog-only, leaving a listed-but-unserved model to the reactive detector
fix(providers): model health reads the catalog only, never a billed completion
All checks were successful
Publish Release / Create Tag & Sync (pull_request) Has been skipped
Details
Publish Release / Publish to crates.io (pull_request) Has been skipped
Details
Publish Release / Publish dry-run (no side effects) (pull_request) Has been skipped
Details
Publish Release / Create GitHub Release (pull_request) Has been skipped
Details
CI / Lint & Audit (pull_request) Successful in 1m6s
Details
CI / Dependency Age Check (pull_request) Successful in 3s
Details
CI / Spell Check (pull_request) Successful in 5s
Details
Commit Lint / Validate PR Title (pull_request) Successful in 3s
Details
CI / Test (pull_request) Successful in 5m0s
Details
55ddd384d0
 
- Removed Interface `ModelAvailability::with_serving_probe` and `probe_serving`, which asked every seat's endpoint for one token on each probe interval
- Fixed Bug where a fleet paid for a completion per seat every five minutes for the life of the process
- Added Redundancy asserting the probe issues no request beyond the catalog fetch
- Fixed Docs describing the proactive check as catalog-only, leaving a listed-but-unserved model to the reactive detector
peersky deleted branch fix/catalog-only-model-health 2026-09-01 07:37:44 +00:00
Sign in to join this conversation.
Reviewers
No reviewers
Labels
Clear labels   
release
Automated release PR

No labels
release
Milestone
Clear milestone
No items
No milestone
Projects
Clear projects
No items
No project
Assignees
Clear assignees
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set

Reference
peeramid-labs/quorum-rs!95
No description provided.