feat: migrate to llm-d-router-gateway chart v0.9.0 and enable KV cache-aware routing - #2144

Open

andyzhangx wants to merge 10 commits into

kaito-project:mainfrom

andyzhangx:feat/migrate-to-llm-d-router-chart

Commits on Aug 1, 2026

feat: migrate from GWIE inferencepool chart to llm-d-router-gateway chart
andyzhangx
committed
fix: gci lint formatting in MRI controller
andyzhangx
committed
fix: restore secure-serving=false in EPP flags
andyzhangx
committed
fix: remove secure-serving=false, use default TLS mode
andyzhangx
committed
fix: reduce EPP resource requests (cpu: 1, memory: 2Gi)
andyzhangx
committed
chore: switch EPP image to mcr.microsoft.com/oss/v2/llm-d/llm-d-router-endpoint-picker:v0.9.0
andyzhangx
committed
chore: bump llm-d-router-gateway chart to v0.9.0

Andy Zhang
authored and
andyzhangx
committed
chore: pin EPP image to v0.9.1 (available on MCR)
andyzhangx
committed
feat: add kv-cache-utilization-scorer to default EPP EndpointPickerConfig
andyzhangx
committed
test: add e2e validation for kv-cache-utilization-scorer in EPP config
andyzhangx
committed