[https://nvbugs/5448767][fix] disable kv cache reuse for disagg pp>1 tests (NVIDIA#7354)

Signed-off-by: Lizhi Zhou <[email protected]>
Signed-off-by: Wangshanshan <[email protected]>

Author: reasonsolo

tests/integration/defs/disaggregated/test_configs/disagg_config_ctxpp2_genpp2.yaml

@@ -14,6 +14,7 @@ context_servers:
   kv_cache_config:
     free_gpu_memory_fraction: 0.2
     enable_partial_reuse: False
+    enable_block_reuse: False
   disable_overlap_scheduler: True
   cache_transceiver_config:
     backend: DEFAULT
@@ -29,6 +30,7 @@ generation_servers:
   kv_cache_config:
     free_gpu_memory_fraction: 0.2
     enable_partial_reuse: False
+    enable_block_reuse: False
   disable_overlap_scheduler: True
   cache_transceiver_config:
     backend: DEFAULT

tests/integration/defs/disaggregated/test_configs/disagg_config_ctxpp2_gentp2.yaml

@@ -14,6 +14,7 @@ context_servers:
   kv_cache_config:
     free_gpu_memory_fraction: 0.2
     enable_partial_reuse: False
+    enable_block_reuse: False
   disable_overlap_scheduler: True
   cache_transceiver_config:
     backend: DEFAULT

tests/integration/defs/disaggregated/test_configs/disagg_config_ctxpp4_genpp4.yaml

@@ -14,6 +14,7 @@ context_servers:
   kv_cache_config:
     free_gpu_memory_fraction: 0.2
     enable_partial_reuse: False
+    enable_block_reuse: False
   disable_overlap_scheduler: True
   cache_transceiver_config:
     backend: DEFAULT
@@ -29,6 +30,7 @@ generation_servers:
   kv_cache_config:
     free_gpu_memory_fraction: 0.2
     enable_partial_reuse: False
+    enable_block_reuse: False
   disable_overlap_scheduler: True
   cache_transceiver_config:
     backend: DEFAULT

tests/integration/defs/disaggregated/test_configs/disagg_config_ctxtp2pp2_gentp2pp2.yaml

@@ -14,6 +14,7 @@ context_servers:
   kv_cache_config:
     free_gpu_memory_fraction: 0.2
     enable_partial_reuse: False
+    enable_block_reuse: False
   disable_overlap_scheduler: True
   cache_transceiver_config:
     backend: DEFAULT
@@ -29,6 +30,7 @@ generation_servers:
   kv_cache_config:
     free_gpu_memory_fraction: 0.2
     enable_partial_reuse: False
+    enable_block_reuse: False
   disable_overlap_scheduler: True
   cache_transceiver_config:
     backend: DEFAULT