Repository navigation
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
An idle Prefill node currently overrides even a strong long-prefix cache hint. For matches of at least 64K characters and 80% of the prompt, compare the cached node's uncached-suffix cost against other nodes using the existing adaptive load threshold. Overloaded cache nodes still shed load; short or low-hit requests keep their current routing.
Validated on main 9640f76: 118 CPU routing, shared-RPC and request-lifecycle tests pass, including 16 new cases for idle/busy peers, overload, threshold boundaries and unavailable cache nodes. This changes routing estimates only; end-to-end TPM impact remains to be measured.