[Refactor]Refactor of vllm_ascend/distributed module (#5719)

### What this PR does / why we need it?
Based on the RFC:https://github.com/vllm-project/vllm-ascend/issues/5604

This PR is a refactoring of vllm_ascend/distributed, moving all
kv_transfer realtaed codes into a dedicated folder, which has already
been done in vLLM

### Does this PR introduce _any_ user-facing change?
NA

### How was this patch tested?


- vLLM version: v0.13.0
- vLLM main:
2f4e6548ef

---------

Signed-off-by: lty <linhebiwen@gmail.com>

docs/source/developer_guide/contribution/multi_node_test.md

@@ -295,7 +295,7 @@ This section assumes that you already have a [Kubernetes](https://kubernetes.io/
     [2025-12-30 11:01:01] INFO multi_node_config.py:348: Resolving cluster IPs via DNS...
     [2025-12-30 11:01:01] INFO multi_node_config.py:212: Node 0 envs: {'VLLM_USE_MODELSCOPE': 'True', 'OMP_PROC_BIND': 'False', 'OMP_NUM_THREADS': '100', 'HCCL_BUFFSIZE': '1024', 'SERVER_PORT': '8080', 'NUMEXPR_MAX_THREADS': '128', 'DISAGGREGATED_PREFILL_PROXY_SCRIPT': 'examples/disaggregated_prefill_v1/load_balance_proxy_server_example.py', 'HCCL_IF_IP': '10.0.0.102', 'HCCL_SOCKET_IFNAME': 'eth0', 'GLOO_SOCKET_IFNAME': 'eth0', 'TP_SOCKET_IFNAME': 'eth0', 'LOCAL_IP': '10.0.0.102', 'NIC_NAME': 'eth0', 'MASTER_IP': '10.0.0.102'}
     [2025-12-30 11:01:01] INFO multi_node_config.py:159: Launching proxy: python examples/disaggregated_prefill_v1/load_balance_proxy_server_example.py --host 10.0.0.102 --port 6000 --prefiller-hosts 10.0.0.102 --prefiller-ports 8080 --decoder-hosts 10.0.0.138 --decoder-ports 8080
-    [2025-12-30 11:01:01] INFO conftest.py:107: Starting server with command: vllm serve vllm-ascend/DeepSeek-V3-W8A8 --host 0.0.0.0 --port 8080 --data-parallel-size 2 --data-parallel-size-local 2 --tensor-parallel-size 8 --seed 1024 --enforce-eager --enable-expert-parallel --max-num-seqs 16 --max-model-len 8192 --max-num-batched-tokens 8192 --quantization ascend --trust-remote-code --no-enable-prefix-caching --gpu-memory-utilization 0.9 --kv-transfer-config {"kv_connector": "MooncakeConnectorV1", "kv_role": "kv_producer", "kv_port": "30000", "engine_id": "0", "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector", "kv_connector_extra_config": {
+    [2025-12-30 11:01:01] INFO conftest.py:107: Starting server with command: vllm serve vllm-ascend/DeepSeek-V3-W8A8 --host 0.0.0.0 --port 8080 --data-parallel-size 2 --data-parallel-size-local 2 --tensor-parallel-size 8 --seed 1024 --enforce-eager --enable-expert-parallel --max-num-seqs 16 --max-model-len 8192 --max-num-batched-tokens 8192 --quantization ascend --trust-remote-code --no-enable-prefix-caching --gpu-memory-utilization 0.9 --kv-transfer-config {"kv_connector": "MooncakeConnectorV1", "kv_role": "kv_producer", "kv_port": "30000", "engine_id": "0", "kv_connector_extra_config": {
             "prefill": {
                     "dp_size": 2,
                     "tp_size": 8

docs/source/tutorials/DeepSeek-V3.1.md

@@ -326,7 +326,6 @@ vllm serve /weights/DeepSeek-V3.1-w8a8-mtp-QuaRot \
   "kv_role": "kv_producer",
   "kv_port": "30000",
   "engine_id": "0",
-  "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
   "kv_connector_extra_config": {
             "prefill": {
                     "dp_size": 2,
@@ -402,7 +401,6 @@ vllm serve /weights/DeepSeek-V3.1-w8a8-mtp-QuaRot \
   "kv_role": "kv_producer",
   "kv_port": "30100",
   "engine_id": "1",
-  "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
   "kv_connector_extra_config": {
             "prefill": {
                     "dp_size": 2,
@@ -480,7 +478,6 @@ vllm serve /weights/DeepSeek-V3.1-w8a8-mtp-QuaRot \
   "kv_role": "kv_consumer",
   "kv_port": "30200",
   "engine_id": "2",
-  "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
   "kv_connector_extra_config": {
             "prefill": {
                     "dp_size": 2,
@@ -558,7 +555,6 @@ vllm serve /weights/DeepSeek-V3.1-w8a8-mtp-QuaRot \
   "kv_role": "kv_consumer",
   "kv_port": "30300",
   "engine_id": "3",
-  "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
   "kv_connector_extra_config": {
             "prefill": {
                     "dp_size": 2,

docs/source/tutorials/DeepSeek-V3.2.md

@@ -316,7 +316,6 @@ Before you start, please
             "kv_role": "kv_producer",
             "kv_port": "30000",
             "engine_id": "0",
-            "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
             "kv_connector_extra_config": {
                         "use_ascend_direct": true,
                         "prefill": {
@@ -391,7 +390,6 @@ Before you start, please
             "kv_role": "kv_producer",
             "kv_port": "30000",
             "engine_id": "0",
-            "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
             "kv_connector_extra_config": {
                         "use_ascend_direct": true,
                         "prefill": {
@@ -469,7 +467,6 @@ Before you start, please
             "kv_role": "kv_consumer",
             "kv_port": "30100",
             "engine_id": "1",
-            "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
             "kv_connector_extra_config": {
                         "use_ascend_direct": true,
                         "prefill": {
@@ -548,7 +545,6 @@ Before you start, please
             "kv_role": "kv_consumer",
             "kv_port": "30100",
             "engine_id": "1",
-            "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
             "kv_connector_extra_config": {
                         "use_ascend_direct": true,
                         "prefill": {

docs/source/tutorials/Qwen3-235B-A22B.md

@@ -450,7 +450,6 @@ vllm serve vllm-ascend/Qwen3-235B-A22B-w8a8 \
 "kv_role": "kv_producer",
 "kv_port": "30000",
 "engine_id": "0",
-"kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
 "kv_connector_extra_config": {
       "use_ascend_direct": true,
       "prefill": {
@@ -516,7 +515,6 @@ vllm serve vllm-ascend/Qwen3-235B-A22B-w8a8 \
 "kv_role": "kv_consumer",
 "kv_port": "30100",
 "engine_id": "1",
-"kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
 "kv_connector_extra_config": {
       "use_ascend_direct": true,
       "prefill": {
@@ -583,7 +581,6 @@ vllm serve vllm-ascend/Qwen3-235B-A22B-w8a8 \
 "kv_role": "kv_consumer",
 "kv_port": "30100",
 "engine_id": "1",
-"kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
 "kv_connector_extra_config": {
       "use_ascend_direct": true,
       "prefill": {

docs/source/tutorials/long_sequence_context_parallel_multi_node.md

@@ -123,7 +123,6 @@ vllm serve /path_to_weight/DeepSeek-V3.1_w8a8mix_mtp \
   "kv_role": "kv_producer",
   "kv_port": "30000",
   "engine_id": "0",
-  "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
   "kv_connector_extra_config": {
             "use_ascend_direct": true,
             "prefill": {
@@ -192,7 +191,6 @@ vllm serve /path_to_weight/DeepSeek-V3.1_w8a8mix_mtp \
   "kv_role": "kv_producer",
   "kv_port": "30000",
   "engine_id": "1",
-  "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
   "kv_connector_extra_config": {
             "use_ascend_direct": true,
             "prefill": {
@@ -259,7 +257,6 @@ vllm serve /path_to_weight/DeepSeek-V3.1_w8a8mix_mtp \
   "kv_role": "kv_consumer",
   "kv_port": "30200",
   "engine_id": "3",
-  "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
   "kv_connector_extra_config": {
             "prefill": {
                     "dp_size": 1,

docs/source/tutorials/pd_disaggregation_mooncake_multi_node.md

@@ -280,7 +280,6 @@ vllm serve /path_to_weight/DeepSeek-r1_w8a8_mtp \
   "kv_role": "kv_producer",
   "kv_port": "30000",
   "engine_id": "0",
-  "kv_connector_module_path": "vllm_ascend.distributed.mooncake_layerwise_connector",
   "kv_connector_extra_config": {
             "prefill": {
                     "dp_size": 2,
@@ -340,7 +339,6 @@ vllm serve /path_to_weight/DeepSeek-r1_w8a8_mtp \
   "kv_role": "kv_producer",
   "kv_port": "30100",
   "engine_id": "1",
-  "kv_connector_module_path": "vllm_ascend.distributed.mooncake_layerwise_connector",
   "kv_connector_extra_config": {
             "prefill": {
                     "dp_size": 2,
@@ -401,7 +399,6 @@ vllm serve /path_to_weight/DeepSeek-r1_w8a8_mtp \
   "kv_role": "kv_consumer",
   "kv_port": "30200",
   "engine_id": "2",
-  "kv_connector_module_path": "vllm_ascend.distributed.mooncake_layerwise_connector",
   "kv_connector_extra_config": {
             "prefill": {
                     "dp_size": 2,
@@ -461,7 +458,6 @@ vllm serve /path_to_weight/DeepSeek-r1_w8a8_mtp \
   "kv_role": "kv_consumer",
   "kv_port": "30200",
   "engine_id": "2",
-  "kv_connector_module_path": "vllm_ascend.distributed.mooncake_layerwise_connector",
   "kv_connector_extra_config": {
             
             "prefill": {
@@ -529,7 +525,6 @@ vllm serve /path_to_weight/DeepSeek-r1_w8a8_mtp \
   "kv_role": "kv_producer",
   "kv_port": "30000",
   "engine_id": "0",
-  "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
   "kv_connector_extra_config": {
             "prefill": {
                     "dp_size": 2,
@@ -589,7 +584,6 @@ vllm serve /path_to_weight/DeepSeek-r1_w8a8_mtp \
   "kv_role": "kv_producer",
   "kv_port": "30100",
   "engine_id": "1",
-  "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
   "kv_connector_extra_config": {
             "prefill": {
                     "dp_size": 2,
@@ -650,7 +644,6 @@ vllm serve /path_to_weight/DeepSeek-r1_w8a8_mtp \
   "kv_role": "kv_consumer",
   "kv_port": "30200",
   "engine_id": "2",
-  "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
   "kv_connector_extra_config": {
             "prefill": {
                     "dp_size": 2,
@@ -710,7 +703,6 @@ vllm serve /path_to_weight/DeepSeek-r1_w8a8_mtp \
   "kv_role": "kv_consumer",
   "kv_port": "30200",
   "engine_id": "2",
-  "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
   "kv_connector_extra_config": {
             "prefill": {
                     "dp_size": 2,

docs/source/tutorials/pd_disaggregation_mooncake_single_node.md

@@ -173,7 +173,6 @@ vllm serve /model/Qwen2.5-VL-7B-Instruct  \
   "kv_role": "kv_producer",
   "kv_port": "30000",
   "engine_id": "0",
-  "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
   "kv_connector_extra_config": {
             "prefill": {
                     "dp_size": 1,
@@ -216,7 +215,6 @@ vllm serve /model/Qwen2.5-VL-7B-Instruct  \
   "kv_role": "kv_consumer",
   "kv_port": "30100",
   "engine_id": "1",
-  "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
   "kv_connector_extra_config": {
             "prefill": {
                     "dp_size": 1,

docs/source/user_guide/deployment_guide/using_volcano_kthena.md

@@ -137,7 +137,7 @@ spec:
                   - "--trust-remote-code"
                   - "--enforce-eager"
                   - "--kv-transfer-config"
-                  - '{"kv_connector":"MooncakeConnectorV1","kv_buffer_device":"npu","kv_role":"kv_producer","kv_parallel_size":1,"kv_port":"20001","engine_id":"0","kv_rank":0,"kv_connector_module_path":"vllm_ascend.distributed.mooncake_connector","kv_connector_extra_config":{"prefill":{"dp_size":2,"tp_size":2},"decode":{"dp_size":2,"tp_size":2}}}'
+                  - '{"kv_connector":"MooncakeConnectorV1","kv_buffer_device":"npu","kv_role":"kv_producer","kv_parallel_size":1,"kv_port":"20001","engine_id":"0","kv_rank":0,"kv_connector_extra_config":{"prefill":{"dp_size":2,"tp_size":2},"decode":{"dp_size":2,"tp_size":2}}}'
                 imagePullPolicy: Always
                 resources:
                   limits:
@@ -240,7 +240,7 @@ spec:
                   - "--no-enable-prefix-caching"
                   - "--enforce-eager"
                   - "--kv-transfer-config"
-                  - '{"kv_connector":"MooncakeConnectorV1","kv_buffer_device":"npu","kv_role":"kv_consumer","kv_parallel_size":1,"kv_port":"20002","engine_id":"1","kv_rank":1,"kv_connector_module_path":"vllm_ascend.distributed.mooncake_connector","kv_connector_extra_config":{"prefill":{"dp_size":2,"tp_size":2},"decode":{"dp_size":2,"tp_size":2}}}'
+                  - '{"kv_connector":"MooncakeConnectorV1","kv_buffer_device":"npu","kv_role":"kv_consumer","kv_parallel_size":1,"kv_port":"20002","engine_id":"1","kv_rank":1,"kv_connector_extra_config":{"prefill":{"dp_size":2,"tp_size":2},"decode":{"dp_size":2,"tp_size":2}}}'
                 imagePullPolicy: Always
                 resources:
                   limits:

docs/source/user_guide/feature_guide/large_scale_ep.md

@@ -163,8 +163,7 @@ vllm serve vllm-ascend/DeepSeek-R1-W8A8 \
       "kv_role": "kv_producer",
       "kv_parallel_size": "1",
       "kv_port": "20001",
-      "engine_id": "0",
-      "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector"
+      "engine_id": "0"
     }'
     --additional-config '{"enable_weight_nz_layout":true,"enable_prefill_optimizations":true}'
 ```
@@ -230,8 +229,7 @@ vllm serve vllm-ascend/DeepSeek-R1-W8A8 \
         "kv_role": "kv_consumer",
         "kv_parallel_size": "1",
         "kv_port": "20001",
-        "engine_id": "0",
-        "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector"
+        "engine_id": "0"
         }' \
     --additional-config '{"enable_weight_nz_layout":true}'
 ```
@@ -435,8 +433,7 @@ In the PD separation scenario, we provide a optimized configuration.
       "kv_role": "kv_producer",
       "kv_parallel_size": "1",
       "kv_port": "20001",
-      "engine_id": "0",
-      "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector"
+      "engine_id": "0"
     }'
 ```
 
@@ -458,8 +455,7 @@ In the PD separation scenario, we provide a optimized configuration.
       "kv_role": "kv_consumer",
       "kv_parallel_size": "1",
       "kv_port": "20001",
-      "engine_id": "0",
-      "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector"
+      "engine_id": "0"
     }'
 ```
 

examples/disaggregated_prefill_v1/mooncake_connector_deployment_guide.md

@@ -55,7 +55,6 @@ vllm serve "/xxxxx/DeepSeek-V2-Lite-Chat" \
   "kv_port": "20001",
   "engine_id": "0",
   "kv_rank": 0,
-  "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
   "kv_connector_extra_config": {
             "prefill": {
                     "dp_size": 2,
@@ -123,7 +122,6 @@ vllm serve "/xxxxx/DeepSeek-V2-Lite-Chat" \
   "kv_port": "20002",
   "engine_id": "1",
   "kv_rank": 1,
-  "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
   "kv_connector_extra_config": {
             "prefill": {
                     "dp_size": 2,

examples/offline_disaggregated_prefill_npu.py

@@ -56,7 +56,6 @@ def run_prefill(prefill_done, process_close):
         kv_role="kv_producer",
         kv_port="30000",
         engine_id="0",
-        kv_connector_module_path="vllm_ascend.distributed.mooncake_connector",
         kv_connector_extra_config={"prefill": {"dp_size": 1, "tp_size": 1}, "decode": {"dp_size": 1, "tp_size": 1}},
     )
     # Set NPU memory utilization to 0.8
@@ -104,7 +103,6 @@ def run_decode(prefill_done):
         kv_role="kv_consumer",
         kv_port="30100",
         engine_id="1",
-        kv_connector_module_path="vllm_ascend.distributed.mooncake_connector",
         kv_connector_extra_config={"prefill": {"dp_size": 1, "tp_size": 1}, "decode": {"dp_size": 1, "tp_size": 1}},
     )
 

tests/e2e/nightly/multi_node/config/DeepSeek-R1-W8A8-EPLB.yaml

@@ -43,7 +43,6 @@ deployment:
           "kv_role": "kv_producer",
           "kv_port": "30000",
           "engine_id": "0",
-          "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
           "kv_connector_extra_config": {
                     "prefill": {
                             "dp_size": 2,
@@ -81,7 +80,6 @@ deployment:
           "kv_role": "kv_producer",
           "kv_port": "30100",
           "engine_id": "1",
-          "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
           "kv_connector_extra_config": {
                     "prefill": {
                             "dp_size": 2,
@@ -120,7 +118,6 @@ deployment:
         "kv_role": "kv_consumer",
         "kv_port": "30200",
         "engine_id": "2",
-        "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
         "kv_connector_extra_config": {
                   "prefill": {
                           "dp_size": 2,
@@ -158,7 +155,6 @@ deployment:
         "kv_role": "kv_consumer",
         "kv_port": "30200",
         "engine_id": "2",
-        "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
         "kv_connector_extra_config": {
                   "prefill": {
                           "dp_size": 2,

tests/e2e/nightly/multi_node/config/DeepSeek-R1-W8A8-longseq.yaml

@@ -46,7 +46,6 @@ deployment:
           "kv_role": "kv_producer",
           "kv_port": "30000",
           "engine_id": "0",
-          "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
           "kv_connector_extra_config": {
                     "prefill": {
                             "dp_size": 1,
@@ -85,7 +84,6 @@ deployment:
         "kv_role": "kv_consumer",
         "kv_port": "30100",
         "engine_id": "1",
-        "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
         "kv_connector_extra_config": {
                   "prefill": {
                           "dp_size": 1,

tests/e2e/nightly/multi_node/config/DeepSeek-R1-W8A8.yaml

@@ -44,7 +44,6 @@ deployment:
           "kv_role": "kv_producer",
           "kv_port": "30000",
           "engine_id": "0",
-          "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
           "kv_connector_extra_config": {
                     "prefill": {
                             "dp_size": 2,
@@ -83,7 +82,6 @@ deployment:
           "kv_role": "kv_producer",
           "kv_port": "30100",
           "engine_id": "1",
-          "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
           "kv_connector_extra_config": {
                     "prefill": {
                             "dp_size": 2,
@@ -124,7 +122,6 @@ deployment:
         "kv_role": "kv_consumer",
         "kv_port": "30200",
         "engine_id": "2",
-        "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
         "kv_connector_extra_config": {
                   "prefill": {
                           "dp_size": 2,
@@ -168,7 +165,6 @@ deployment:
         "kv_role": "kv_consumer",
         "kv_port": "30200",
         "engine_id": "2",
-        "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
         "kv_connector_extra_config": {
                   "prefill": {
                           "dp_size": 2,

tests/e2e/nightly/multi_node/config/DeepSeek-V3.yaml

@@ -53,7 +53,6 @@ deployment:
         "kv_role": "kv_producer",
         "kv_port": "30000",
         "engine_id": "0",
-        "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
         "kv_connector_extra_config": {
                   "prefill": {
                           "dp_size": 2,
@@ -88,7 +87,6 @@ deployment:
         "kv_role": "kv_consumer",
         "kv_port": "30200",
         "engine_id": "1",
-        "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
         "kv_connector_extra_config": {
                   "prefill": {
                           "dp_size": 2,

tests/e2e/nightly/multi_node/config/Qwen3-235B-W8A8-EPLB.yaml

@@ -39,7 +39,6 @@ deployment:
         "kv_role": "kv_producer",
         "kv_port": "30000",
         "engine_id": "0",
-        "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
         "kv_connector_extra_config": {
                   "prefill": {
                           "dp_size": 2,
@@ -76,7 +75,6 @@ deployment:
         "kv_role": "kv_consumer",
         "kv_port": "30200",
         "engine_id": "1",
-        "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
         "kv_connector_extra_config": {
                   "prefill": {
                           "dp_size": 2,

tests/e2e/nightly/multi_node/config/Qwen3-235B-W8A8-longseq.yaml

@@ -43,7 +43,6 @@ deployment:
         "kv_role": "kv_producer",
         "kv_port": "30000",
         "engine_id": "0",
-        "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
         "kv_connector_extra_config": {
                   "prefill": {
                           "dp_size": 1,
@@ -83,7 +82,6 @@ deployment:
         "kv_role": "kv_consumer",
         "kv_port": "30100",
         "engine_id": "1",
-        "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
         "kv_connector_extra_config": {
                   "prefill": {
                           "dp_size": 1,

tests/e2e/nightly/multi_node/config/Qwen3-235B-W8A8.yaml

@@ -39,7 +39,6 @@ deployment:
         "kv_role": "kv_producer",
         "kv_port": "30000",
         "engine_id": "0",
-        "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
         "kv_connector_extra_config": {
                   "prefill": {
                           "dp_size": 2,
@@ -74,7 +73,6 @@ deployment:
         "kv_role": "kv_consumer",
         "kv_port": "30200",
         "engine_id": "1",
-        "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
         "kv_connector_extra_config": {
                   "prefill": {
                           "dp_size": 2,

tests/e2e/nightly/multi_node/config/Qwen3-235B-disagg-pd.yaml

@@ -44,7 +44,6 @@ deployment:
         "kv_role": "kv_producer",
         "kv_port": "30000",
         "engine_id": "0",
-        "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
         "kv_connector_extra_config": {
                   "use_ascend_direct": true,
                   "prefill": {
@@ -84,7 +83,6 @@ deployment:
         "kv_role": "kv_consumer",
         "kv_port": "30100",
         "engine_id": "1",
-        "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
         "kv_connector_extra_config": {
                   "use_ascend_direct": true,
                   "prefill": {

tests/e2e/nightly/multi_node/config/Qwen3-VL-235B-disagg-pd.yaml

@@ -39,7 +39,6 @@ deployment:
       "kv_role": "kv_producer",
       "kv_port": "30000",
       "engine_id": "0",
-      "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
       "kv_connector_extra_config": {
                 "prefill": {
                         "dp_size": 2,
@@ -73,7 +72,6 @@ deployment:
       "kv_role": "kv_consumer",
       "kv_port": "30200",
       "engine_id": "1",
-      "kv_connector_module_path": "vllm_ascend.distributed.mooncake_connector",
       "kv_connector_extra_config": {
                 "prefill": {
                         "dp_size": 2,

tests/ut/distributed/mooncake/test_config_data.py

@@ -10,7 +10,7 @@ fake_store = types.ModuleType("mooncake.store")
 fake_store.ReplicateConfig = MagicMock()  # type: ignore[attr-defined]
 sys.modules["mooncake.store"] = fake_store
 
-from vllm_ascend.distributed.kvpool.backend.mooncake_backend import (  # noqa: E402
+from vllm_ascend.distributed.kv_transfer.kv_pool.ascend_store.backend.mooncake_backend import (  # noqa: E402
     _convert_to_bytes, _parse_global_segment_size)
 
 

tests/ut/distributed/test_communicator.py

@@ -4,7 +4,8 @@ from unittest.mock import MagicMock, patch
 import torch
 import torch.distributed as dist
 
-from vllm_ascend.distributed.communicator import NPUCommunicator
+from vllm_ascend.distributed.device_communicators.npu_communicator import \
+    NPUCommunicator
 
 
 class TestNPUCommunicator(unittest.TestCase):

tests/ut/kv_connector/test_mooncake_connector.py

@@ -23,21 +23,24 @@ _mock_pp_group = MagicMock(rank_in_group=0, world_size=1)
 _mock_tp_group = MagicMock(rank_in_group=0, world_size=4)
 _mock_pcp_group = MagicMock(rank_in_group=0, world_size=1)
 _mock_dcp_group = MagicMock(rank_in_group=0, world_size=1)
-patch('vllm_ascend.distributed.mooncake_connector.get_pp_group',
-      return_value=_mock_pp_group).start()
-patch('vllm_ascend.distributed.mooncake_connector.get_tp_group',
-      return_value=_mock_tp_group).start()
 patch(
-    'vllm_ascend.distributed.mooncake_connector.get_tensor_model_parallel_world_size',
+    'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.get_pp_group',
+    return_value=_mock_pp_group).start()
+patch(
+    'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.get_tp_group',
+    return_value=_mock_tp_group).start()
+patch(
+    'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.get_tensor_model_parallel_world_size',
     return_value=4).start()
 patch(
-    'vllm_ascend.distributed.mooncake_connector.get_tensor_model_parallel_rank',
+    'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.get_tensor_model_parallel_rank',
     return_value=0).start()
-patch('vllm_ascend.distributed.mooncake_connector.get_pcp_group',
-      return_value=_mock_pcp_group).start()
+patch(
+    'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.get_pcp_group',
+    return_value=_mock_pcp_group).start()
 patch('vllm.distributed.parallel_state._DCP', _mock_dcp_group).start()
 
-from vllm_ascend.distributed.mooncake_connector import (  # noqa: E402
+from vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector import (  # noqa: E402
     KVCacheRecvingThread, KVCacheSendingThread, KVCacheTaskTracker,
     KVConnectorRole, MooncakeAgentMetadata, MooncakeConnector,
     MooncakeConnectorMetadata, MooncakeConnectorScheduler,
@@ -81,7 +84,8 @@ class TestGetAndClearFinishedSingleRequests(unittest.TestCase):
         self.assertSetEqual(result, {"req_1", "req_2", "req_3"})
         self.assertEqual(len(self.tracker.finished_requests), 0)
 
-    @patch("vllm_ascend.distributed.mooncake_connector.logger")
+    @patch(
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.logger")
     def test_concurrent_access(self, mock_logger):
         from concurrent.futures import ThreadPoolExecutor
         self.tracker.finished_requests = {"req_1", "req_2"}
@@ -307,8 +311,12 @@ class TestSocketManagement(unittest.TestCase):
         self.thread.remote_sockets = defaultdict(deque)
         self.thread.remote_poller = MagicMock()
 
-    @patch('vllm_ascend.distributed.mooncake_connector.zmq.Context')
-    @patch('vllm_ascend.distributed.mooncake_connector.make_zmq_socket')
+    @patch(
+        'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.zmq.Context'
+    )
+    @patch(
+        'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.make_zmq_socket'
+    )
     def test_get_remote_socket(self, mock_make_socket, mock_context):
         mock_sock = MagicMock()
         mock_make_socket.return_value = mock_sock
@@ -402,7 +410,7 @@ class TestCoreFunctionality(unittest.TestCase):
     @patch.object(KVCacheRecvingThread, '_get_remote_metadata')
     def test_transfer_kv_cache(self, mock_get_meta):
         with patch(
-                'vllm_ascend.distributed.mooncake_connector.get_ascend_config'
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.get_ascend_config'
         ) as mock_config:
             mock_config.return_value.enable_kv_nz = False
             self.thread.kv_caches_base_addr["remote_engine"] = {
@@ -456,8 +464,12 @@ class TestMetadataHandling(unittest.TestCase):
             kv_caches_base_addr=[0x3000, 0x4000],
             num_blocks=2)
 
-    @patch('vllm_ascend.distributed.mooncake_connector.ensure_zmq_send')
-    @patch('vllm_ascend.distributed.mooncake_connector.ensure_zmq_recv')
+    @patch(
+        'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.ensure_zmq_send'
+    )
+    @patch(
+        'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.ensure_zmq_recv'
+    )
     def test_get_remote_metadata_success(self, mock_recv, mock_send):
         mock_recv.return_value = msgspec.msgpack.encode(self.test_metadata)
 
@@ -479,9 +491,12 @@ class TestMetadataHandling(unittest.TestCase):
                 self.thread.kv_caches_base_addr["remote_engine"][5555],
                 [0x3000, 0x4000])
 
-    @patch('vllm_ascend.distributed.mooncake_connector.ensure_zmq_send')
-    @patch('vllm_ascend.distributed.mooncake_connector.ensure_zmq_recv',
-           side_effect=Exception("Network error"))
+    @patch(
+        'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.ensure_zmq_send'
+    )
+    @patch(
+        'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.ensure_zmq_recv',
+        side_effect=Exception("Network error"))
     def test_get_remote_metadata_failure(self, mock_recv, mock_send):
         with patch.object(self.thread, '_get_remote_socket') as mock_get_socket, \
                 patch.object(self.thread, '_return_remote_socket') as mock_return_socket:
@@ -694,10 +709,10 @@ class TestMooncakeConnectorSchedulerMatchedTokens(unittest.TestCase):
     def setUp(self):
         config = MockVllmConfig()
         self.p1 = patch(
-            'vllm_ascend.distributed.mooncake_connector.init_ascend_config',
+            'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.init_ascend_config',
             new=MagicMock())
         self.p2 = patch(
-            'vllm_ascend.distributed.mooncake_connector.get_ascend_config',
+            'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.get_ascend_config',
             new=MagicMock(return_value=MagicMock()))
         self.p1.start()
         self.p2.start()
@@ -775,9 +790,9 @@ class TestMooncakeConnectorForScheduler(unittest.TestCase):
     def test_scheduler_role(self):
         config = MockVllmConfig()
         with patch(
-                'vllm_ascend.distributed.mooncake_connector.init_ascend_config'
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.init_ascend_config'
         ), patch(
-                'vllm_ascend.distributed.mooncake_connector.get_ascend_config',
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.get_ascend_config',
                 return_value=MagicMock()):
             connector = MooncakeConnector(config, KVConnectorRole.SCHEDULER)
         self.assertIsNotNone(connector.connector_scheduler)
@@ -787,9 +802,9 @@ class TestMooncakeConnectorForScheduler(unittest.TestCase):
     def test_scheduler_methods(self, mock_method):
         config = MockVllmConfig()
         with patch(
-                'vllm_ascend.distributed.mooncake_connector.init_ascend_config'
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.init_ascend_config'
         ), patch(
-                'vllm_ascend.distributed.mooncake_connector.get_ascend_config',
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.get_ascend_config',
                 return_value=MagicMock()):
             connector = MooncakeConnector(config, KVConnectorRole.SCHEDULER)
         request = MockRequest("req1")
@@ -819,9 +834,9 @@ class TestMooncakeConnector(unittest.TestCase):
 
     def test_scheduler_initialization(self):
         with patch(
-                'vllm_ascend.distributed.mooncake_connector.init_ascend_config'
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.init_ascend_config'
         ), patch(
-                'vllm_ascend.distributed.mooncake_connector.get_ascend_config',
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.get_ascend_config',
                 return_value=MagicMock()):
             connector = MooncakeConnector(self.config,
                                           KVConnectorRole.SCHEDULER)
@@ -831,9 +846,9 @@ class TestMooncakeConnector(unittest.TestCase):
     @patch.object(MooncakeConnectorScheduler, "get_num_new_matched_tokens")
     def test_get_num_new_matched_tokens(self, mock_method):
         with patch(
-                'vllm_ascend.distributed.mooncake_connector.init_ascend_config'
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.init_ascend_config'
         ), patch(
-                'vllm_ascend.distributed.mooncake_connector.get_ascend_config',
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.get_ascend_config',
                 return_value=MagicMock()):
             connector = MooncakeConnector(self.config,
                                           KVConnectorRole.SCHEDULER)
@@ -844,9 +859,9 @@ class TestMooncakeConnector(unittest.TestCase):
     @patch.object(MooncakeConnectorScheduler, "update_state_after_alloc")
     def test_update_state_after_alloc(self, mock_method):
         with patch(
-                'vllm_ascend.distributed.mooncake_connector.init_ascend_config'
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.init_ascend_config'
         ), patch(
-                'vllm_ascend.distributed.mooncake_connector.get_ascend_config',
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.get_ascend_config',
                 return_value=MagicMock()):
             connector = MooncakeConnector(self.config,
                                           KVConnectorRole.SCHEDULER)
@@ -858,9 +873,9 @@ class TestMooncakeConnector(unittest.TestCase):
     @patch.object(MooncakeConnectorScheduler, "build_connector_meta")
     def test_build_connector_meta(self, mock_method):
         with patch(
-                'vllm_ascend.distributed.mooncake_connector.init_ascend_config'
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.init_ascend_config'
         ), patch(
-                'vllm_ascend.distributed.mooncake_connector.get_ascend_config',
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.get_ascend_config',
                 return_value=MagicMock()):
             connector = MooncakeConnector(self.config,
                                           KVConnectorRole.SCHEDULER)
@@ -871,9 +886,9 @@ class TestMooncakeConnector(unittest.TestCase):
     @patch.object(MooncakeConnectorScheduler, "request_finished")
     def test_request_finished(self, mock_method):
         with patch(
-                'vllm_ascend.distributed.mooncake_connector.init_ascend_config'
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.init_ascend_config'
         ), patch(
-                'vllm_ascend.distributed.mooncake_connector.get_ascend_config',
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.get_ascend_config',
                 return_value=MagicMock()):
             connector = MooncakeConnector(self.config,
                                           KVConnectorRole.SCHEDULER)
@@ -887,9 +902,9 @@ class TestMooncakeConnectorScheduler(unittest.TestCase):
     def setUp(self):
         self.config = MockVllmConfig()
         with patch(
-                'vllm_ascend.distributed.mooncake_connector.init_ascend_config'
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.init_ascend_config'
         ), patch(
-                'vllm_ascend.distributed.mooncake_connector.get_ascend_config',
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.get_ascend_config',
                 return_value=MagicMock()):
             self.scheduler = MooncakeConnectorScheduler(
                 self.config, "test_engine")
@@ -965,20 +980,24 @@ class TestUtils(unittest.TestCase):
             with zmq_ctx("INVALID", "tcp://127.0.0.1:5555"):
                 pass
 
-    @patch("vllm_ascend.distributed.mooncake_connector.make_zmq_socket")
+    @patch(
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.make_zmq_socket"
+    )
     def test_zmq_ctx_ok(self, mock_make_socket):
         mock_socket = MagicMock()
         mock_make_socket.return_value = mock_socket
         with zmq_ctx(zmq.REQ, "tcp://localhost:1234") as s:  # type: ignore
             self.assertEqual(s, mock_socket)
 
-    @patch("vllm_ascend.distributed.mooncake_connector.logger")
+    @patch(
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.logger")
     def test_ensure_zmq_send_success(self, mock_logger):
         mock_socket = MagicMock()
         ensure_zmq_send(mock_socket, b"hello")
         mock_socket.send.assert_called_once_with(b"hello")
 
-    @patch("vllm_ascend.distributed.mooncake_connector.logger")
+    @patch(
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.logger")
     def test_ensure_zmq_send_retry_and_fail(self, mock_logger):
         mock_socket = MagicMock()
         mock_socket.send.side_effect = zmq.ZMQError(  # type: ignore
@@ -987,7 +1006,8 @@ class TestUtils(unittest.TestCase):
             ensure_zmq_send(mock_socket, b"hello", max_retries=2)
         self.assertEqual(mock_socket.send.call_count, 2)
 
-    @patch("vllm_ascend.distributed.mooncake_connector.logger")
+    @patch(
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.logger")
     def test_ensure_zmq_recv_success(self, mock_logger):
         mock_socket = MagicMock()
         mock_socket.recv.return_value = b"response"
@@ -998,7 +1018,8 @@ class TestUtils(unittest.TestCase):
         data = ensure_zmq_recv(mock_socket, mock_poller)
         self.assertEqual(data, b"response")
 
-    @patch("vllm_ascend.distributed.mooncake_connector.logger")
+    @patch(
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.logger")
     def test_ensure_zmq_recv_timeout_and_fail(self, mock_logger):
         mock_socket = MagicMock()
         mock_poller = MagicMock()
@@ -1106,35 +1127,40 @@ class TestMooncakeConnectorWorker(unittest.TestCase):
             patch('torch.Tensor.data_ptr', return_value=0x1000),
             patch('math.prod', return_value=128),
             patch(
-                'vllm_ascend.distributed.mooncake_connector.get_tensor_model_parallel_rank',
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.get_tensor_model_parallel_rank',
                 mock_get_tensor_model_parallel_rank),
-            patch('vllm_ascend.distributed.mooncake_connector.get_tp_group',
-                  mock_get_tp_group),
-            patch('vllm_ascend.distributed.mooncake_connector.get_pp_group',
-                  return_value=_mock_pp_group),
-            patch('vllm_ascend.distributed.mooncake_connector.get_ip',
-                  mock_get_ip),
             patch(
-                'vllm_ascend.distributed.mooncake_connector.string_to_int64_hash',
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.get_tp_group',
+                mock_get_tp_group),
+            patch(
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.get_pp_group',
+                return_value=_mock_pp_group),
+            patch(
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.get_ip',
+                mock_get_ip),
+            patch(
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.string_to_int64_hash',
                 mock_string_to_int64_hash),
             patch(
-                'vllm_ascend.distributed.mooncake_connector.global_te.get_transfer_engine',
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.global_te.get_transfer_engine',
                 return_value=self.mock_transfer_engine),
             patch(
-                'vllm_ascend.distributed.mooncake_connector.global_te.register_buffer',
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.global_te.register_buffer',
                 return_value=None),
             patch(
-                'vllm_ascend.distributed.mooncake_connector.KVCacheSendingThread',
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.KVCacheSendingThread',
                 MagicMock()),
             patch(
-                'vllm_ascend.distributed.mooncake_connector.KVCacheRecvingThread',
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.KVCacheRecvingThread',
                 MagicMock()),
-            patch('vllm_ascend.distributed.mooncake_connector.logger',
-                  MagicMock()),
-            patch('vllm_ascend.distributed.mooncake_connector.threading.Event',
-                  MagicMock()),
             patch(
-                'vllm_ascend.distributed.mooncake_connector.get_ascend_config',
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.logger',
+                MagicMock()),
+            patch(
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.threading.Event',
+                MagicMock()),
+            patch(
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.get_ascend_config',
                 return_value=MagicMock()),
         ]
 
@@ -1186,7 +1212,8 @@ class TestMooncakeConnectorWorker(unittest.TestCase):
         def get_tp_rank(prefill_tp_size: int, prefill_pp_size: int,
                         decode_tp_size: int, num_kv_heads: int,
                         tp_num_need_pulls: int, is_deepseek_mla: bool):
-            with patch('vllm_ascend.distributed.mooncake_connector.get_ascend_config',
+            with patch(
+                    'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector.get_ascend_config',
                     return_value=MagicMock()), \
                 patch.object(self.vllm_config.kv_transfer_config, 'get_from_extra_config',
                             side_effect=lambda k, d=None: {

tests/ut/kv_connector/test_mooncake_layerwise_connector.py

@@ -15,7 +15,7 @@ fake_engine = types.ModuleType("mooncake.engine")
 fake_engine.TransferEngine = MagicMock()  # type: ignore[attr-defined]
 sys.modules["mooncake.engine"] = fake_engine
 
-from vllm_ascend.distributed.mooncake_layerwise_connector import (  # noqa: E402
+from vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector import (  # noqa: E402
     KVCacheRecvingLayerThread, KVCacheSendingLayerThread, KVConnectorRole,
     MooncakeAgentMetadata, MooncakeLayerwiseConnector,
     MooncakeLayerwiseConnectorMetadata, MooncakeLayerwiseConnectorScheduler,
@@ -81,19 +81,20 @@ class TestKVCacheSendingLayerThread(unittest.TestCase):
             chunk_finish=False)
 
     @patch(
-        "vllm_ascend.distributed.mooncake_layerwise_connector.npu_stream_switch",
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.npu_stream_switch",
         side_effect=lambda *_args, **_kwargs: contextlib.nullcontext())
     @patch(
-        "vllm_ascend.distributed.mooncake_layerwise_connector.torch.Tensor.data_ptr",
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.torch.Tensor.data_ptr",
         autospec=True,
         return_value=0x200000)
-    @patch("vllm_ascend.distributed.mooncake_layerwise_connector.align_memory",
-           side_effect=lambda x, _align: x)
     @patch(
-        "vllm_ascend.distributed.mooncake_layerwise_connector.torch.npu.synchronize"
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.align_memory",
+        side_effect=lambda x, _align: x)
+    @patch(
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.torch.npu.synchronize"
     )
     @patch(
-        "vllm_ascend.distributed.mooncake_layerwise_connector.group_concurrent_contiguous"
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.group_concurrent_contiguous"
     )
     def test_transfer_pd_gt1_uses_buffers_and_calls_engine(
             self, mock_group, _mock_sync, _mock_align, _mock_dataptr,
@@ -171,10 +172,10 @@ class TestKVCacheSendingLayerThread(unittest.TestCase):
         self.engine.batch_transfer_sync_write.assert_not_called()
 
     @patch(
-        "vllm_ascend.distributed.mooncake_layerwise_connector.group_concurrent_contiguous",
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.group_concurrent_contiguous",
         side_effect=group_concurrent_contiguous)
     @patch(
-        "vllm_ascend.distributed.mooncake_layerwise_connector.torch.npu.synchronize"
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.torch.npu.synchronize"
     )
     def test_callback_invoked_on_final_layer(self, _mock_sync, _mock_group):
 
@@ -250,21 +251,27 @@ class TestKVCacheRecvingLayerThread(unittest.TestCase):
             self.assertNotIn("reqX", th.task_tracker)
             self.assertIn("reqX", th.done_requests)
 
-    @patch("vllm_ascend.distributed.mooncake_layerwise_connector.logger")
-    @patch("vllm_ascend.distributed.mooncake_layerwise_connector.get_ip",
-           return_value="127.0.0.1")
     @patch(
-        "vllm_ascend.distributed.mooncake_layerwise_connector.make_zmq_socket")
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.logger"
+    )
     @patch(
-        "vllm_ascend.distributed.mooncake_layerwise_connector.make_zmq_path",
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.get_ip",
+        return_value="127.0.0.1")
+    @patch(
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.make_zmq_socket"
+    )
+    @patch(
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.make_zmq_path",
         side_effect=lambda proto, host, port: f"{proto}://{host}:{port}")
     @patch(
-        "vllm_ascend.distributed.mooncake_layerwise_connector.msgspec.msgpack.Decoder"
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.msgspec.msgpack.Decoder"
     )
     @patch(
-        "vllm_ascend.distributed.mooncake_layerwise_connector.msgspec.msgpack.Encoder"
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.msgspec.msgpack.Encoder"
+    )
+    @patch(
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.zmq_ctx"
     )
-    @patch("vllm_ascend.distributed.mooncake_layerwise_connector.zmq_ctx")
     def test_run_loop_handles_meta_done_invalid_unexpected_and_ack(
             self, mock_zmq_ctx, mock_Encoder, mock_Decoder, _mock_make_path,
             _mock_make_sock, _mock_get_ip, mock_logger):
@@ -330,16 +337,21 @@ class TestKVCacheRecvingLayerThread(unittest.TestCase):
         finished = th.get_and_clear_finished_requests()
         self.assertIn("reqA", finished)
 
-    @patch("vllm_ascend.distributed.mooncake_layerwise_connector.logger")
-    @patch("vllm_ascend.distributed.mooncake_layerwise_connector.get_ip",
-           return_value="127.0.0.1")
     @patch(
-        "vllm_ascend.distributed.mooncake_layerwise_connector.msgspec.msgpack.Decoder"
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.logger"
     )
     @patch(
-        "vllm_ascend.distributed.mooncake_layerwise_connector.msgspec.msgpack.Encoder"
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.get_ip",
+        return_value="127.0.0.1")
+    @patch(
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.msgspec.msgpack.Decoder"
+    )
+    @patch(
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.msgspec.msgpack.Encoder"
+    )
+    @patch(
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.zmq_ctx"
     )
-    @patch("vllm_ascend.distributed.mooncake_layerwise_connector.zmq_ctx")
     def test_run_loop_pd_head_ratio_gt1_requires_multiple_done(
             self, mock_zmq_ctx, mock_Encoder, mock_Decoder, _mock_get_ip,
             _mock_logger):
@@ -623,7 +635,7 @@ class TestMooncakeLayerwiseConnectorScheduler_More(unittest.TestCase):
         self.assertEqual(len(meta.requests), 0)
 
     @patch(
-        "vllm_ascend.distributed.mooncake_layerwise_connector.group_concurrent_contiguous"
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.group_concurrent_contiguous"
     )
     def test_build_connector_meta_emits_when_tokens_reach_total(
             self, mock_group_concurrent_contiguous):
@@ -707,20 +719,25 @@ class TestHelperFunctions(unittest.TestCase):
                 pass
 
     @patch(
-        "vllm_ascend.distributed.mooncake_layerwise_connector.make_zmq_socket")
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.make_zmq_socket"
+    )
     def test_zmq_ctx_ok(self, mock_make_socket):
         mock_socket = MagicMock()
         mock_make_socket.return_value = mock_socket
         with zmq_ctx(zmq.REQ, "tcp://localhost:1234") as s:  # type: ignore
             self.assertEqual(s, mock_socket)
 
-    @patch("vllm_ascend.distributed.mooncake_layerwise_connector.logger")
+    @patch(
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.logger"
+    )
     def test_ensure_zmq_send_success(self, _):
         mock_socket = MagicMock()
         ensure_zmq_send(mock_socket, b"hello")
         mock_socket.send.assert_called_once_with(b"hello")
 
-    @patch("vllm_ascend.distributed.mooncake_layerwise_connector.logger")
+    @patch(
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.logger"
+    )
     def test_ensure_zmq_send_retry_and_fail(self, _):
         mock_socket = MagicMock()
         mock_socket.send.side_effect = zmq.ZMQError(  # type: ignore
@@ -729,7 +746,9 @@ class TestHelperFunctions(unittest.TestCase):
             ensure_zmq_send(mock_socket, b"hello", max_retries=2)
         self.assertEqual(mock_socket.send.call_count, 2)
 
-    @patch("vllm_ascend.distributed.mooncake_layerwise_connector.logger")
+    @patch(
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.logger"
+    )
     def test_ensure_zmq_recv_success(self, _):
         mock_socket = MagicMock()
         mock_socket.recv.return_value = b"response"
@@ -740,7 +759,9 @@ class TestHelperFunctions(unittest.TestCase):
         data = ensure_zmq_recv(mock_socket, mock_poller)
         self.assertEqual(data, b"response")
 
-    @patch("vllm_ascend.distributed.mooncake_layerwise_connector.logger")
+    @patch(
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.logger"
+    )
     def test_ensure_zmq_recv_timeout_and_fail(self, _):
         mock_socket = MagicMock()
         mock_poller = MagicMock()
@@ -849,37 +870,37 @@ class TestMooncakeLayerwiseConnectorWorker(unittest.TestCase):
             patch('math.prod', return_value=128),
             patch('random.Random'),
             patch(
-                'vllm_ascend.distributed.mooncake_layerwise_connector.get_tensor_model_parallel_rank',
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.get_tensor_model_parallel_rank',
                 return_value=0),
             patch(
-                'vllm_ascend.distributed.mooncake_layerwise_connector.get_tp_group',
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.get_tp_group',
                 return_value=None),
             patch(
-                'vllm_ascend.distributed.mooncake_layerwise_connector.get_ip',
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.get_ip',
                 return_value="127.0.0.1"),
             patch(
-                'vllm_ascend.distributed.mooncake_layerwise_connector.string_to_int64_hash',
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.string_to_int64_hash',
                 side_effect=lambda s: hash(s)),
             patch(
-                'vllm_ascend.distributed.mooncake_layerwise_connector.global_te.get_transfer_engine',
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.global_te.get_transfer_engine',
                 return_value=self.mock_transfer_engine),
             patch(
-                'vllm_ascend.distributed.mooncake_layerwise_connector.global_te.register_buffer',
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.global_te.register_buffer',
                 return_value=None),
             patch(
-                'vllm_ascend.distributed.mooncake_layerwise_connector.KVCacheSendingLayerThread',
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.KVCacheSendingLayerThread',
                 MagicMock()),
             patch(
-                'vllm_ascend.distributed.mooncake_layerwise_connector.KVCacheRecvingLayerThread',
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.KVCacheRecvingLayerThread',
                 MagicMock()),
             patch(
-                'vllm_ascend.distributed.mooncake_layerwise_connector.logger',
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.logger',
                 MagicMock()),
             patch(
-                'vllm_ascend.distributed.mooncake_layerwise_connector.threading.Event',
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.threading.Event',
                 MagicMock()),
             patch(
-                'vllm_ascend.distributed.mooncake_layerwise_connector.get_ascend_config',
+                'vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector.get_ascend_config',
                 return_value=SimpleNamespace(pd_tp_ratio=1,
                                              num_head_replica=1,
                                              pd_head_ratio=1)),

tests/ut/kv_connector/utils.py

@@ -79,8 +79,7 @@ def create_vllm_config(
     )
     kv_transfer_config = KVTransferConfig(
         kv_connector="MooncakeConnectorV1",
-        kv_role="kv_both",
-        kv_connector_module_path="vllm_ascend.distributed.mooncake_connector")
+        kv_role="kv_both")
     return VllmConfig(scheduler_config=scheduler_config,
                       model_config=model_config,
                       cache_config=cache_config,

tests/ut/test_platform.py

@@ -461,7 +461,7 @@ class TestNPUPlatform(TestBase):
     def test_get_device_communicator_cls_returns_correct_value(self):
         self.assertEqual(
             self.platform.get_device_communicator_cls(),
-            "vllm_ascend.distributed.communicator.NPUCommunicator",
+            "vllm_ascend.distributed.device_communicators.npu_communicator.NPUCommunicator",
         )
 
     def test_is_pin_memory_available_returns_true(self):

vllm_ascend/__init__.py

@@ -23,7 +23,7 @@ def register():
 
 
 def register_connector():
-    from vllm_ascend.distributed import register_connector
+    from vllm_ascend.distributed.kv_transfer import register_connector
     register_connector()
 
 

vllm_ascend/distributed/__init__.py

@@ -1,44 +0,0 @@
-#
-# Copyright (c) 2025 Huawei Technologies Co., Ltd. All Rights Reserved.
-# This file is a part of the vllm-ascend project.
-#
-# Licensed under the Apache License, Version 2.0 (the "License");
-# you may not use this file except in compliance with the License.
-# You may obtain a copy of the License at
-#
-#     http://www.apache.org/licenses/LICENSE-2.0
-#
-# Unless required by applicable law or agreed to in writing, software
-# distributed under the License is distributed on an "AS IS" BASIS,
-# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-# See the License for the specific language governing permissions and
-# limitations under the License.
-#
-
-from vllm.distributed.kv_transfer.kv_connector.factory import \
-    KVConnectorFactory
-
-
-def register_connector():
-    KVConnectorFactory.register_connector(
-        "MooncakeConnectorV1", "vllm_ascend.distributed.mooncake_connector",
-        "MooncakeConnector")
-
-    KVConnectorFactory.register_connector(
-        "MooncakeConnectorStoreV1",
-        "vllm_ascend.distributed.kvpool.ascend_store_connector",
-        "AscendStoreConnector")
-
-    KVConnectorFactory.register_connector(
-        "AscendStoreConnector",
-        "vllm_ascend.distributed.kvpool.ascend_store_connector",
-        "AscendStoreConnector")
-
-    KVConnectorFactory.register_connector(
-        "MooncakeLayerwiseConnector",
-        "vllm_ascend.distributed.mooncake_layerwise_connector",
-        "MooncakeLayerwiseConnector")
-
-    KVConnectorFactory.register_connector(
-        "UCMConnector", "vllm_ascend.distributed.ucm_connector",
-        "UCMConnectorV1")

vllm_ascend/distributed/kv_transfer/__init__.py

@@ -0,0 +1,45 @@
+#
+# Copyright (c) 2025 Huawei Technologies Co., Ltd. All Rights Reserved.
+# This file is a part of the vllm-ascend project.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+#
+
+from vllm.distributed.kv_transfer.kv_connector.factory import \
+    KVConnectorFactory
+
+
+def register_connector():
+    KVConnectorFactory.register_connector(
+        "MooncakeConnectorV1",
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector",
+        "MooncakeConnector")
+
+    KVConnectorFactory.register_connector(
+        "MooncakeConnectorStoreV1",
+        "vllm_ascend.distributed.kv_transfer.kv_pool.ascend_store.ascend_store_connector",
+        "AscendStoreConnector")
+
+    KVConnectorFactory.register_connector(
+        "AscendStoreConnector",
+        "vllm_ascend.distributed.kv_transfer.kv_pool.ascend_store.ascend_store_connector",
+        "AscendStoreConnector")
+
+    KVConnectorFactory.register_connector(
+        "MooncakeLayerwiseConnector",
+        "vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_layerwise_connector",
+        "MooncakeLayerwiseConnector")
+
+    KVConnectorFactory.register_connector(
+        "UCMConnector", "vllm_ascend.distributed.kv_transfer.ucm_connector",
+        "UCMConnectorV1")

vllm_ascend/distributed/kv_transfer/kv_p2p/mooncake_connector.py

@@ -41,8 +41,8 @@ from vllm.v1.kv_cache_interface import KVCacheConfig
 from vllm.v1.request import RequestStatus
 
 from vllm_ascend.ascend_config import get_ascend_config, init_ascend_config
-from vllm_ascend.distributed.mooncake_transfer_engine import global_te
-from vllm_ascend.distributed.utils import get_transfer_timeout_value
+from vllm_ascend.distributed.kv_transfer.utils.mooncake_transfer_engine import global_te
+from vllm_ascend.distributed.kv_transfer.utils.utils import get_transfer_timeout_value
 from vllm_ascend.utils import is_vl_model
 
 if TYPE_CHECKING:

vllm_ascend/distributed/kv_transfer/kv_p2p/mooncake_layerwise_connector.py

@@ -32,11 +32,12 @@ from vllm.v1.core.sched.output import SchedulerOutput
 from vllm.v1.kv_cache_interface import KVCacheConfig
 
 from vllm_ascend.ascend_config import get_ascend_config
-from vllm_ascend.distributed.mooncake_connector import GET_META_MSG
-from vllm_ascend.distributed.mooncake_transfer_engine import global_te
-from vllm_ascend.distributed.utils import (align_memory,
-                                           get_transfer_timeout_value,
-                                           kv_alltoall_and_rearrange)
+from vllm_ascend.distributed.kv_transfer.kv_p2p.mooncake_connector import \
+    GET_META_MSG
+from vllm_ascend.distributed.kv_transfer.utils.mooncake_transfer_engine import \
+    global_te
+from vllm_ascend.distributed.kv_transfer.utils.utils import (
+    align_memory, get_transfer_timeout_value, kv_alltoall_and_rearrange)
 from vllm_ascend.utils import npu_stream_switch
 
 if TYPE_CHECKING:

vllm_ascend/distributed/kv_transfer/kv_pool/ascend_store/ascend_store_connector.py

@@ -16,9 +16,10 @@ from vllm.v1.kv_cache_interface import KVCacheConfig
 from vllm.v1.request import Request
 from vllm.v1.serial_utils import MsgpackDecoder
 
-from vllm_ascend.distributed.kvpool.pool_scheduler import (
+from vllm_ascend.distributed.kv_transfer.kv_pool.ascend_store.pool_scheduler import (
     KVPoolScheduler, get_zmq_rpc_path_lookup)
-from vllm_ascend.distributed.kvpool.pool_worker import KVPoolWorker
+from vllm_ascend.distributed.kv_transfer.kv_pool.ascend_store.pool_worker import \
+    KVPoolWorker
 
 
 class AscendStoreConnector(KVConnectorBase_V1):

vllm_ascend/distributed/kv_transfer/kv_pool/ascend_store/backend/memcache_backend.py

@@ -5,7 +5,8 @@ import torch
 from vllm.config import ParallelConfig
 from vllm.logger import logger
 
-from vllm_ascend.distributed.kvpool.backend.backend import Backend
+from vllm_ascend.distributed.kv_transfer.kv_pool.ascend_store.backend.backend import \
+    Backend
 from vllm_ascend.utils import AscendDeviceType, get_ascend_device_type
 
 

vllm_ascend/distributed/kv_transfer/kv_pool/ascend_store/backend/mooncake_backend.py

@@ -11,8 +11,10 @@ from vllm.config import ParallelConfig
 from vllm.logger import logger
 from vllm.utils.network_utils import get_ip
 
-from vllm_ascend.distributed.kvpool.backend.backend import Backend
-from vllm_ascend.distributed.mooncake_transfer_engine import global_te
+from vllm_ascend.distributed.kv_transfer.kv_pool.ascend_store.backend.backend import \
+    Backend
+from vllm_ascend.distributed.kv_transfer.utils.mooncake_transfer_engine import \
+    global_te
 
 DEFAULT_GLOBAL_SEGMENT_SIZE = 3355443200  # 3.125 GiB
 DEFAULT_LOCAL_BUFFER_SIZE = 1073741824  # 1.0 GiB

vllm_ascend/distributed/kv_transfer/kv_pool/ascend_store/kv_transfer.py

@@ -7,10 +7,11 @@ from typing import Any
 import torch
 from vllm.logger import logger
 
-from vllm_ascend.distributed.kvpool.backend.backend import Backend
+from vllm_ascend.distributed.kv_transfer.kv_pool.ascend_store.backend.backend import \
+    Backend
 
 # isort: off
-from vllm_ascend.distributed.kvpool.config_data import (
+from vllm_ascend.distributed.kv_transfer.kv_pool.ascend_store.config_data import (
     ChunkedTokenDatabase,
     LasyerMultiBlockReqMeta,
     ReqMeta,

vllm_ascend/distributed/kv_transfer/kv_pool/ascend_store/metadata.py

@@ -14,7 +14,7 @@ from vllm.utils.network_utils import make_zmq_socket
 from vllm.utils.torch_utils import get_dtype_size
 from vllm.v1.kv_cache_interface import AttentionSpec, MLAAttentionSpec
 
-from vllm_ascend.distributed.cpu_offload_manager.cpu_kv_cache_manager import \
+from vllm_ascend.distributed.kv_transfer.kv_pool.cpu_offload.cpu_kv_cache_manager import \
     CPUKVCacheManager
 
 

vllm_ascend/distributed/kv_transfer/kv_pool/ascend_store/pool_scheduler.py

@@ -13,7 +13,7 @@ from vllm.v1.core.sched.output import SchedulerOutput
 from vllm.v1.request import Request
 from vllm.v1.serial_utils import MsgpackEncoder
 
-from vllm_ascend.distributed.kvpool.config_data import (
+from vllm_ascend.distributed.kv_transfer.kv_pool.ascend_store.config_data import (
     AscendConnectorMetadata, LoadSpec, ReqMeta, RequestTracker)
 
 

vllm_ascend/distributed/kv_transfer/kv_pool/ascend_store/pool_worker.py

@@ -11,15 +11,16 @@ from vllm.distributed import (get_decode_context_model_parallel_rank,
 from vllm.logger import logger
 from vllm.v1.core.kv_cache_utils import BlockHash
 
-from vllm_ascend.distributed.kvpool.backend.backend import Backend
-from vllm_ascend.distributed.kvpool.backend.memcache_backend import \
+from vllm_ascend.distributed.kv_transfer.kv_pool.ascend_store.backend.backend import \
+    Backend
+from vllm_ascend.distributed.kv_transfer.kv_pool.ascend_store.backend.memcache_backend import \
     MemcacheBackend
-from vllm_ascend.distributed.kvpool.backend.mooncake_backend import \
+from vllm_ascend.distributed.kv_transfer.kv_pool.ascend_store.backend.mooncake_backend import \
     MooncakeBackend
-from vllm_ascend.distributed.kvpool.config_data import (
+from vllm_ascend.distributed.kv_transfer.kv_pool.ascend_store.config_data import (
     AscendConnectorMetadata, ChunkedTokenDatabase, KeyMetadata,
     LasyerMultiBlockReqMeta, ReqMeta)
-from vllm_ascend.distributed.kvpool.kv_transfer import (
+from vllm_ascend.distributed.kv_transfer.kv_pool.ascend_store.kv_transfer import (
     KVCacheStoreLayerRecvingThread, KVCacheStoreLayerSendingThread,
     KVCacheStoreRecvingThread, KVCacheStoreSendingThread, KVTransferThread)
 

vllm_ascend/distributed/kv_transfer/kv_pool/cpu_offload/cpu_offload_connector.py

@@ -24,7 +24,7 @@ from vllm.v1.core.sched.output import SchedulerOutput
 from vllm.v1.kv_cache_interface import (FullAttentionSpec, KVCacheSpec,
                                         MambaSpec, MLAAttentionSpec)
 
-from vllm_ascend.distributed.cpu_offload_manager.metadata import (
+from vllm_ascend.distributed.kv_transfer.kv_pool.ascend_store.metadata import (
     MetadataServer, MetadataServerProc, MLAConfig)
 
 if TYPE_CHECKING:

vllm_ascend/distributed/kv_transfer/utils/utils.py

@@ -0,0 +1,61 @@
+import os
+
+import torch
+import torch.distributed as dist
+
+from vllm_ascend.distributed.parallel_state import get_p_tp_group
+
+
+def kv_alltoall_and_rearrange(pd_tp_ratio: int, key: torch.Tensor,
+                              value: torch.TensorType):
+    if pd_tp_ratio <= 1:
+        return None, None
+    elif key is None or value is None:
+        raise ValueError("key or value is None")
+    k_output = alltoall_and_rearrange(pd_tp_ratio, key)
+    v_output = alltoall_and_rearrange(pd_tp_ratio, value)
+    return k_output, v_output
+
+
+def alltoall_and_rearrange(tp_ratio: int, input_tensor: torch.Tensor):
+    num_kv_heads = input_tensor.size(1)
+    output_tensor = torch.zeros_like(input_tensor)
+    dist.all_to_all_single(output_tensor,
+                           input_tensor,
+                           group=get_p_tp_group().device_group)
+    input_tensor = 0
+    result = rearrange_output(output_tensor, tp_ratio, num_kv_heads)
+    output_tensor = 0
+    return result
+
+
+def rearrange_output(base_output: torch.Tensor, cut_num: int,
+                     num_kv_heads: int):
+    size_0 = base_output.size(0)
+    if size_0 % cut_num != 0:
+        raise ValueError(
+            f"The size of dim 0 [{size_0}] must be divisible by the cut_num [{cut_num}]"
+        )
+    chunk_size = size_0 // cut_num
+    reshaped = base_output.view(cut_num, chunk_size, -1)
+    transposed = reshaped.transpose(0, 1)
+    return transposed.contiguous().view(size_0, num_kv_heads, -1)
+
+
+def align_memory(tensor: torch.Tensor, alignment: int) -> torch.Tensor:
+    data_ptr = tensor.data_ptr()
+    aligned_addr = (data_ptr + alignment - 1) // alignment * alignment
+    offset = (aligned_addr - data_ptr) // tensor.element_size()
+    return tensor[int(offset):]
+
+
+def get_transfer_timeout_value():
+    ascend_transfer_timeout = os.getenv("ASCEND_TRANSFER_TIMEOUT", "")
+    if len(ascend_transfer_timeout) > 0:
+        return int(ascend_transfer_timeout)
+    hccl_rdma_timeout = int(os.getenv('HCCL_RDMA_TIMEOUT',
+                                      '20'))  # type: ignore
+    hccl_rdma_retry_cnt = int(os.getenv('HCCL_RDMA_RETRY_CNT',
+                                        '7'))  # type: ignore
+    return int((4.096 * (2**hccl_rdma_timeout)) * hccl_rdma_retry_cnt // 1000 +
+               3000)

vllm_ascend/distributed/kvpool/backend/__init__.py

@@ -1 +0,0 @@
-

vllm_ascend/distributed/utils.py

@@ -1,4 +1,3 @@
-import os
 from typing import Optional
 
 import torch
@@ -6,63 +5,7 @@ import torch.distributed as dist
 from vllm.distributed.parallel_state import GroupCoordinator, get_dp_group
 from vllm.forward_context import get_forward_context
 
-from vllm_ascend.distributed.parallel_state import (get_fc3_quant_x_group,
-                                                    get_p_tp_group)
-
-
-def kv_alltoall_and_rearrange(pd_tp_ratio: int, key: torch.Tensor,
-                              value: torch.TensorType):
-    if pd_tp_ratio <= 1:
-        return None, None
-    elif key is None or value is None:
-        raise ValueError("key or value is None")
-    k_output = alltoall_and_rearrange(pd_tp_ratio, key)
-    v_output = alltoall_and_rearrange(pd_tp_ratio, value)
-    return k_output, v_output
-
-
-def alltoall_and_rearrange(tp_ratio: int, input_tensor: torch.Tensor):
-    num_kv_heads = input_tensor.size(1)
-    output_tensor = torch.zeros_like(input_tensor)
-    dist.all_to_all_single(output_tensor,
-                           input_tensor,
-                           group=get_p_tp_group().device_group)
-    input_tensor = 0
-    result = rearrange_output(output_tensor, tp_ratio, num_kv_heads)
-    output_tensor = 0
-    return result
-
-
-def rearrange_output(base_output: torch.Tensor, cut_num: int,
-                     num_kv_heads: int):
-    size_0 = base_output.size(0)
-    if size_0 % cut_num != 0:
-        raise ValueError(
-            f"The size of dim 0 [{size_0}] must be divisible by the cut_num [{cut_num}]"
-        )
-    chunk_size = size_0 // cut_num
-    reshaped = base_output.view(cut_num, chunk_size, -1)
-    transposed = reshaped.transpose(0, 1)
-    return transposed.contiguous().view(size_0, num_kv_heads, -1)
-
-
-def align_memory(tensor: torch.Tensor, alignment: int) -> torch.Tensor:
-    data_ptr = tensor.data_ptr()
-    aligned_addr = (data_ptr + alignment - 1) // alignment * alignment
-    offset = (aligned_addr - data_ptr) // tensor.element_size()
-    return tensor[int(offset):]
-
-
-def get_transfer_timeout_value():
-    ascend_transfer_timeout = os.getenv("ASCEND_TRANSFER_TIMEOUT", "")
-    if len(ascend_transfer_timeout) > 0:
-        return int(ascend_transfer_timeout)
-    hccl_rdma_timeout = int(os.getenv('HCCL_RDMA_TIMEOUT',
-                                      '20'))  # type: ignore
-    hccl_rdma_retry_cnt = int(os.getenv('HCCL_RDMA_RETRY_CNT',
-                                        '7'))  # type: ignore
-    return int((4.096 * (2**hccl_rdma_timeout)) * hccl_rdma_retry_cnt // 1000 +
-               3000)
+from vllm_ascend.distributed.parallel_state import get_fc3_quant_x_group
 
 
 def fc3_all_gather_and_maybe_unpad_impl(x: torch.Tensor, ) -> torch.Tensor:
@@ -90,6 +33,7 @@ def fc3_all_gather_and_maybe_unpad_impl(x: torch.Tensor, ) -> torch.Tensor:
             result[offset:offset + num_tokens_dp] = x[idx, :num_tokens_dp]
             offset += num_tokens_dp
         x = result
+
     return x
 
 

vllm_ascend/patch/worker/patch_distributed.py

@@ -23,7 +23,8 @@ from torch.distributed import Backend
 from vllm.distributed.parallel_state import (GroupCoordinator,
                                              _get_unique_name, _register_group)
 
-from vllm_ascend.distributed.communicator import NPUCommunicator
+from vllm_ascend.distributed.device_communicators.npu_communicator import \
+    NPUCommunicator
 from vllm_ascend.utils import create_hccl_pg_options
 
 

vllm_ascend/platform.py

@@ -411,7 +411,7 @@ class NPUPlatform(Platform):
 
     @classmethod
     def get_device_communicator_cls(cls) -> str:
-        return "vllm_ascend.distributed.communicator.NPUCommunicator"
+        return "vllm_ascend.distributed.device_communicators.npu_communicator.NPUCommunicator"
 
     @classmethod
     def is_pin_memory_available(cls):