Unverified50% confidenceFactExact time
Cache-aware routing preferentially sends requests to replicas that already have the same prefix in cache; when the cache is hit, prefill is almost zero, and first-token latency drops significantly.
1
Sources
50%
Confidence
Long-term
Relevance
10/12/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedPrefix caching is extremely sensitive to input stability; as long as any single token in the prefix is modified or inserted, the cache is invalidated and all subsequent content must be recomputed.71% similarUnverifiedPrefix caching skips redundant computation by reusing previously computed KV cache for identical prefixes from prior requests, but any formatting change, order perturbation, or insertion of content in the middle will invalidate the cache.71% similarUnverifiedIn primary-replica replication, asynchronous replication mode does not require the primary to wait for replica acknowledgment before committing transactions, resulting in replication lag; semi-synchronous mode requires at least one replica to acknowledge before returning success, reducing the risk of data loss but increasing write latency.70% similarUnverifiedIn the cache reuse mechanism, once the request prefix is modified, the cache for all content after it becomes invalid; modifying tools or other parts of the request will still affect reuse.70% similarUnverifiedv2.1.269 fixes an issue where the prompt cache was partially invalidated when a response was truncated upon hitting the output token limit and then automatically continued.69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1011565API
curl https://kongchang.com/api/v1/knowledge/claims/1011565MCP
get_claim(id=1011565)