fix: config should use new precise-prefix-cache-scorer

- we have rename prefix-cache-scorer to precise-prefix-cache-scorer in 0.3.0, configs
  need migrate from the old one to the new one with spec.
  - rename plugin name
  - remove parameters.autoTune and parameters.mode: cache_tracking and
    lruCapacityPerServer
  - move hashBlockSize, maxPrefixBlocksToMatch under indexrConfig
- for config using food-review keep old prefix-cache-scorer
- keep pd-epp-config and sim-pd-epp-config with prefix-cache-scorer as
  KV and PD need both be enabled which is not done yet

Signed-off-by: Wen Zhou <wenzhou@redhat.com>

Author: zdtsw

deploy/config/sim-epp-kvcache-config.yaml

@@ -5,15 +5,12 @@ kind: EndpointPickerConfig
 plugins:
 - type: precise-prefix-cache-scorer
   parameters:
-    mode: cache_tracking
-    tokenProcessorConfig:
-      blockSize: 16                         # must match vLLM block size if not default (16)
-      hashSeed: "42"                        # must match PYTHONHASHSEED in vLLM pods
     kvEventsConfig:
       zmqEndpoint: tcp://0.0.0.0:5557
     indexerConfig:
-      prefixStoreConfig:
-        blockSize: 16 
+      tokenProcessorConfig:
+        blockSize: 16                         # must match vLLM block size if not default (16)
+        hashSeed: "42"                        # must match PYTHONHASHSEED in vLLM pods
       tokenizersPoolConfig:
         modelName: TinyLlama/TinyLlama-1.1B-Chat-v1.0  # replace value to use different model for tokenizer loading
         hf:

docs/architecture.md

@@ -161,11 +161,13 @@ A complete configuration might look like this:
 apiVersion: inference.networking.x-k8s.io/v1alpha1
 kind: EndpointPickerConfig
 plugins:
-- type: prefix-cache-scorer
+- type: precise-prefix-cache-scorer
   parameters:
-    hashBlockSize: 5
-    maxPrefixBlocksToMatch: 256
-    lruCapacityPerServer: 31250
+    indexerConfig:
+      tokenProcessorConfig:
+        blockSize: 5
+      kvBlockIndexConfig:
+        maxPrefixBlocksToMatch: 256
 - type: decode-filter
 - type: max-score-picker
 - type: single-profile-handler
@@ -174,7 +176,7 @@ schedulingProfiles:
   plugins:
   - pluginRef: decode-filter
   - pluginRef: max-score-picker
-  - pluginRef: prefix-cache-scorer
+  - pluginRef: precise-prefix-cache-scorer
     weight: 50
 ```
 
@@ -465,11 +467,13 @@ Example configuration:
 
 ```yaml
 plugins:
-  - type: prefix-cache-scorer
+  - type: precise-prefix-cache-scorer
     parameters:
-      hashBlockSize: 5
-      maxPrefixBlocksToMatch: 256
-      lruCapacityPerServer: 31250
+      indexerConfig:
+        tokenProcessorConfig:
+          blockSize: 5
+        kvBlockIndexConfig:
+          maxPrefixBlocksToMatch: 256
   - type: no-hit-lru-scorer
     parameters:
       lruSize: 2048
@@ -481,7 +485,7 @@ schedulingProfiles:
     plugins:
       - pluginRef: decode-filter
       - pluginRef: max-score-picker
-      - pluginRef: prefix-cache-scorer
+      - pluginRef: precise-prefix-cache-scorer
         weight: 2
       - pluginRef: no-hit-lru-scorer
         weight: 1
@@ -502,11 +506,13 @@ apiVersion: inference.networking.x-k8s.io/v1alpha1
 kind: EndpointPickerConfig
 plugins:
 - type: prefill-header-handler
-- type: prefix-cache-scorer
+- type: precise-prefix-cache-scorer
   parameters:
-    hashBlockSize: 5
-    maxPrefixBlocksToMatch: 256
-    lruCapacityPerServer: 31250
+    indexerConfig:
+      tokenProcessorConfig:
+        blockSize: 5
+      kvBlockIndexConfig:
+        maxPrefixBlocksToMatch: 256
 - type: prefill-filter
 - type: decode-filter
 - type: max-score-picker
@@ -519,13 +525,13 @@ schedulingProfiles:
   plugins:
   - pluginRef: prefill-filter
   - pluginRef: max-score-picker
-  - pluginRef: prefix-cache-scorer
+  - pluginRef: precise-prefix-cache-scorer
     weight: 50
 - name: decode
   plugins:
   - pluginRef: decode-filter
   - pluginRef: max-score-picker
-  - pluginRef: prefix-cache-scorer
+  - pluginRef: precise-prefix-cache-scorer
     weight: 50
 ```