apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2026-04-25 09:57:51 +08:00
Code Issues Actions 9 Packages Projects Releases Wiki Activity
5,093 Commits 40 Branches 17 Tags
d8cdda86cb022f210fdeafadaee054e26f7acf87
Commit Graph

105 Commits

Author SHA1 Message Date
kevin 7707be8384 [Feature][KVCache] Implement Cache Manager V1 with GPU + CPU Cache Support (1/n) (#7097) 
* [Feature][KVCache] Support cache manager v1 architecture

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Update cache manager and related modules

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* chore: update cache_manager and related modules

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: add node to evictable set in complete_swap_to_device

When a node transitions from SWAP_TO_DEVICE to DEVICE via
complete_swap_to_device, it was not being added to the
_evictable_device set. This caused nodes with ref_count=0 to
become "orphaned" - not appearing in any evictable set despite
having cache_status=DEVICE.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* feat: update cache manager v1 and related modules

- Add new cache_manager.py with cache management functionality
- Add radix_tree.py for prefix caching
- Update block_pool.py and metadata.py
- Update request.py and resource_manager_v1.py for scheduling
- Update gpu_model_runner.py for GPU model execution

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* feat(cache): add cache controller v1 implementation

- Add CacheController class for cache management
- Update config.py with cache related configurations
- Refactor gpu_model_runner.py for improved cache handling

* feat(cache_manager): update cache manager v1

* fix(cache_manager): 修复 swap_cache H2D/D2H 方向的 block_ids 逻辑并清理 ForwardMeta

## Motivation

修复 swap_cache_optimized.cu 中 H2D 方向时 src/dst block_ids 使用错误的问题，
并清理 ForwardMeta 中已废弃的 cache_controller 字段。

## Modifications

- fix: swap_cache_optimized.cu 中根据 D2H 模板参数正确选取 src/dst block_ids，
  修复 H2D 方向 src/dst 倒置 bug（同时修复 SwapCachePerLayerImpl 和 SwapCacheAllLayersBatchImpl）
- refactor: cache_manager/v1/__init__.py 将 LayerSwapTimeoutError 导入从
  cache_controller 改为 cache_utils（正确来源）
- refactor: ForwardMeta 移除废弃的 cache_controller 字段
- refactor: gpu_model_runner.py 移除对应的 cache_controller 赋值语句
- test: 新增 tests/cache_manager/v1/test_swap_cache_ops.py 单元测试

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* feat(cache_manager): refactor cache manager v1 and optimize swap ops

## Motivation

对 cache manager v1 进行重构和优化，精简代码结构，提升可维护性。

## Modifications

- 重构 transfer_manager.py，大幅精简代码逻辑
- 优化 swap_cache_optimized.cu GPU 算子实现
- 调整 cache_manager.py、cache_controller.py 逻辑，修复 free_device_blocks 方法缺失问题
- 更新 block_pool.py、cache_utils.py、metadata.py、radix_tree.py
- 精简 gpu_model_runner.py、forward_meta.py、attention.py 中相关调用
- 更新对应单元测试（test_cache_controller、test_swap_cache_ops、test_transfer_manager）
- 调整 config.py 中相关配置项

* [KVCache][MTP] 支持 cache_manager_v1 下的 MTP KV Cache 初始化及多模态 hash

## Motivation

在 enable_cache_manager_v1 路径下，MTP（speculative decode）的 KV Cache 需要由
CacheController 统一管理，以复用 swap/transfer 能力，同时修复多模态场景下 block
hash 未携带 multimodal extra_keys 的问题。

## Modifications

- `cache_controller.py`
  - 新增 `initialize_mtp_kv_cache`：通过 CacheController 初始化 MTP KV Cache，
    并将其注册到 cache_kvs_map，使 transfer_manager 自动覆盖 MTP 层
  - `initialize_host_cache` 中的 num_layers 改为包含 MTP 额外 cache 层数，保证
    Host Cache 也为 MTP 分配足够空间
  - `_free_gpu_cache` 改名为 `free_gpu_cache`（对外可调用）

- `cache_utils.py`
  - 新增 `get_block_hash_extra_keys`：提取单个 block 内的多模态 hash 信息，
    对齐 PrefixCacheManager 的 multimodal extra_keys 逻辑
  - `get_request_block_hasher` 中在 hash_block_tokens 时携带 extra_keys，
    修复多模态场景 prefix cache 命中率不准的问题

- `spec_decode/mtp.py`
  - `update_mtp_block_num` 新增 `skip_cache_init` 参数，避免 v1 cache manager
    路径下重复初始化 MTP KV Cache

- `gpu_model_runner.py`
  - `initialize_kv_cache(v1)` 路径：在主模型 cache 初始化后，调用
    `cache_controller.initialize_mtp_kv_cache` 完成 MTP cache 创建
  - `clear_cache` / `wakeup` / `reset` 等路径：respect `enable_cache_manager_v1`
    标志，跳过重复的 proposer.initialize_kv_cache 调用

## Usage or Command

```bash
# 启动支持 MTP + cache_manager_v1 的推理服务（示例）
bash run.sh
```

* fix(cache_manager): multi-GPU fix, mm hash boundary fix, and remove batch ops

1. Fix CuPy stream/event creation for multi-GPU: wrap all stream operations
   with cp.cuda.Device(device_id) context to ensure streams/events are bound
   to the correct device, preventing cross-device errors in multi-GPU setups.

2. Remove cudaSetDevice from SwapCacheAllLayers (handled by cupy context now).

3. Remove swap_cache_all_layers_batch op: simplified the implementation by
   removing the batch upload variant; all-layer transfers now use the standard
   swap_cache_all_layers with cupy device context.

4. Fix mm hash boundary comparison in get_block_hash_extra_keys: change
   strict less-than (<) to less-than-or-equal (<=) so that multimodal items
   ending exactly at block start are correctly excluded.

5. Extract config fields to KVCacheBase: model_config, cache_config,
   quant_config, parallel_config are now set in the base class __init__ to
   avoid duplication in CacheController and CacheManager subclasses.

6. Translate metadata.py docstrings from Chinese to English for broader
   contributor accessibility.

7. Add test_cache_utils.py: comprehensive unit tests for
   get_block_hash_extra_keys covering all boundary and overlap scenarios.

8. Expand test suite: test_request.py cache fields tests, test_radix_tree.py
   backup candidate tests, test_transfer_manager.py and test_cache_manager.py
   multi-GPU and concurrent operation tests.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* [BugFix][KVCache] fix List import and move write_policy normalization to CacheManager

## Motivation

修复两处问题：
1. `fastdeploy/engine/request.py` 中 `List` 未导入导致 pre-commit F821 报错
2. `write_policy` 归一化逻辑（`write_through` → `write_through_selective`）不应放在 `FDConfig`，移至 `CacheManager.__init__` 中，使其只影响 Cache Manager V1 的内部逻辑

## Modifications

- `fastdeploy/engine/request.py`: 在 `typing` 导入中补充 `List`，删除重复的 `CacheSwapMetadata` TYPE_CHECKING 导入，修复 F821/F811
- `fastdeploy/config.py`: 删除 `write_policy` 归一化逻辑
- `fastdeploy/cache_manager/v1/cache_manager.py`: 将归一化逻辑移入 `CacheManager.__init__`

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* [BugFix][KVCache] fix pre-commit code style issues

## Motivation

修复 CI pre-commit 代码风格检查失败问题。

## Modifications

- `fastdeploy/engine/common_engine.py`: black 格式化
- `fastdeploy/worker/worker_process.py`: black 格式化 + isort 修复
- `fastdeploy/cache_manager/v1/storage/__init__.py`: isort 修复
- `fastdeploy/worker/gpu_worker.py`: isort 修复

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* [Feature][KVCache] update cache_manager_v1 modules

## Motivation

更新 Cache Manager V1 相关模块，完善版权信息、改进模块结构与可维护性。

## Modifications

- `fastdeploy/cache_manager/v1/` 系列模块：补充版权 header，优化代码结构
- `fastdeploy/config.py`：配置项更新
- `fastdeploy/engine/sched/resource_manager_v1.py`：调度相关更新

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* [Feature][KVCache] add BatchRequest.from_tasks and refactor worker task parsing

## Motivation

将 worker_process 中重复的 task 解析逻辑收敛到 BatchRequest，减少代码冗余，提升可维护性。

## Modifications

- `fastdeploy/engine/request.py`：新增 `BatchRequest.from_tasks()` 类方法，统一将 task_queue 任务分类为推理请求和控制请求
- `fastdeploy/worker/worker_process.py`：使用 `BatchRequest.from_tasks()` 替代内联解析逻辑，并修复重复的 control_reqs 处理块

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* [Feature][KVCache] add NUMA affinity for host cache and skip swap cache tests

## Motivation

优化 Host cache 内存分配的 NUMA 亲和性，减少跨 NUMA 访问延迟；
同时跳过 swap cache ops 测试（当前环境不支持）。

## Modifications

- `fastdeploy/cache_manager/v1/cache_controller.py`：
  - 新增 `_get_numa_node_for_gpu()` 方法，通过 nvidia-smi 或 sysfs 获取 GPU 对应的 NUMA 节点
  - 新增 `_bind_to_closest_numa_node()` 方法，绑定当前线程到 GPU 最近的 NUMA 节点
  - 在 `initialize_host_cache()` 中调用 NUMA 绑定，优化 H2D 传输性能
- `tests/cache_manager/v1/test_swap_cache_ops.py`：跳过所有测试类（`TestSwapCacheAllLayersCorrectness`、`TestSwapCacheAllLayersPerformance`、`TestSwapCacheRandomBlockIndices`）

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* [BugFix][KVCache] fix unittest failures for cache_manager_v1

三个单测因接口变更或 Mock 方式问题导致失败，需修复。

- tests/distributed/chunked_moe.py：`setup_model_runner` 使用 `__new__` 跳过 `__init__`，补加 `enable_cache_manager_v1 = False`，修复 `AttributeError`
- tests/engine/test_resource_manager.py：`PrefixCacheManager` 为局部导入，`patch` 路径改为定义位置 `fastdeploy.cache_manager.prefix_cache_manager.PrefixCacheManager`
- tests/v1/test_resource_manager_v1.py：`_trigger_preempt` 第四参数已由 `list` 改为 `BatchRequest`，更新测试传参和断言

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* [BugFix][KVCache] remove debug logging code

## Modifications

- fastdeploy/engine/request.py：删除调试用 logger 及 prompt_hashes 中的 debug 日志
- fastdeploy/worker/worker_process.py：删除 __main__ 中的调试 import 和 print 语句

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* [BugFix][KVCache] fix cupy device id caching and pickle for _match_result

## Motivation

修复两个 bug：
1. `transfer_manager.py` 中每次调用 `cp.cuda.runtime.getDevice()` 存在隐患，应在初始化时缓存为实例变量，保证后续操作使用一致的设备 ID。
2. `request.py` 的 `__getstate__` 未跳过 `_match_result`，该字段包含 BlockNode 树的父子循环引用，pickle 时会触发 `RecursionError`；同时补充 `__setstate__` 确保 unpickle 后字段恢复为安全默认值。

## Modifications

- `transfer_manager.py`：初始化时调用 `cp.cuda.runtime.getDevice()` 并缓存到 `self._cupy_device_id`，后续 `with cp.cuda.Device(...)` 和日志均使用该缓存值。
- `request.py`：
  - `__getstate__` 中将 `_match_result` 加入跳过集合 `_SKIP_KEYS`，避免循环引用导致 pickle 失败。
  - 新增 `__setstate__`，unpickle 后将 `_block_hasher` 和 `_match_result` 恢复为 `None`。

## Usage or Command

* fix(test): fix unit test errors for _trigger_preempt and wakeup with MTP

- Use BatchRequest instead of list in test_trigger_preempt_records_tasks
- Add missing enable_cache_manager_v1 attr in TestSleepWakeupBehavior._make_runner

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* [BugFix][KVCache] fix gpu_free_block_list returning wrong block IDs

## Motivation

`gpu_free_block_list` 的兼容 property 中误用了 `list(range(N))`，
将 `available_blocks()` 的返回值当作整数传给 `range()`，
导致返回 `[0, 1, ..., N-1]` 的假列表，而非真实的空闲 block ID。

## Modifications

- `cache_manager/v1/cache_manager.py`：将 `list(range(self._device_pool.available_blocks()))` 改为 `list(self._device_pool.available_blocks())`

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* [BugFix][KVCache] 修复 gpu_free_block_list 返回 int 导致 TypeError

## Motivation

gpu_free_block_list 属性中调用 BlockPool.available_blocks()，
该方法返回 int（空闲块数量），用 list() 包装 int 会触发
TypeError: 'int' object is not iterable。

## Modifications

将 list(self._device_pool.available_blocks()) 改为
list(self._device_pool._free_blocks)，直接返回空闲块索引列表。

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* [KVCache][CacheManager] 适配 V1 CacheManager 的 pause/sleep/free_cache 操作

## Motivation

V1 CacheManager 引入了新的 reset_cache() 接口，pause 和 sleep 操作需要适配，
同时 free_cache 需要支持可选的 clear_storage 参数。

## Modifications

- cache_controller.py: free_cache 新增 clear_storage 参数（默认 False），
  仅当 clear_storage=True 时才调用 _clear_storage()，避免不必要的 storage 清空
- common_engine.py: pause 和 sleep 操作中，当 ENABLE_V1_KVCACHE_MANAGER 时
  使用 cache_manager.reset_cache() 替代旧的 reset() 和 pause_transfer 逻辑
- gpu_model_runner.py: sleep 时仅在非 V1 cache manager 下执行 MTP cache 清除

## Usage or Command

# 启动服务（V1 CacheManager）
python -m fastdeploy.entrypoints.openai.api_server \
  --enable-v1-kvcache-manager \
  ...

* [BugFix][KVCache] fix missing enable_cache_manager_v1 in test mocks and remove unused select_blocks_for_backup

- Remove unused `select_blocks_for_backup` method from radix_tree.py
- Fix `match_prefix` default param `skip_storage=True` and log order in cache_manager.py
- Sync test_gpu_model_runner.py with upstream/develop (add TestInsertTasksV1SplitwiseSuffix)
- Add `enable_cache_manager_v1=False` to all mock runners to fix AttributeError in CI

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* [BugFix][KVCache] simplify _free_blocks in ResourceManagerV1 for non-v1 path

Remove redundant prefix_caching branch in else path; always call
recycle_gpu_blocks with full block_tables for non-cache-manager-v1 case.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* [KVCache][Optimization][BugFix] fix and optimize block_pool, cache_manager, transfer_manager, request

## Motivation

修复 cache_manager v1 中若干代码质量问题，提升性能并消除潜在的类型不一致 Bug。

## Modifications

1. **block_pool.py**：`BlockPool.allocate` 将逐个 pop 循环替换为切片 + 批量 set.update，消除 Python 循环开销，O(n) → O(k)（C 层批量操作）
2. **cache_manager.py**：`match_prefix` 在 prefix caching 关闭时提前 return 前写入空 `MatchResult()`，避免调用方解引用 `_match_result=None` 崩溃
3. **transfer_manager.py**：`_build_device_layer_indices` 在 `_cache_kvs_map` 为空时也重置四个层索引列表，防止残留旧 tensor 被 swap 算子使用
4. **request.py**：`BatchRequest.append_swap_metadata` / `append_evict_metadata` 构造 `CacheSwapMetadata` 时将 `src_type`/`dst_type` 从字符串改为 `CacheLevel` 枚举，与字段类型声明一致；补充 `CacheLevel` 导入；`match_result` 属性返回类型标注修正为 `Optional[MatchResult]`
5. **resource_manager_v1.py**：`_allocate_gpu_blocks` 日志从 `INFO` 降级为 `DEBUG`，消除高频调度路径的日志噪音
6. **tests/engine/test_request.py**：同步更新 `src_type`/`dst_type` 断言为 `CacheLevel` 枚举值，补充 `CacheLevel` 导入

## Usage or Command

单元测试：
```bash
source .venv/py310/bin/activate
cd baidu/FastDeploy
python -m pytest tests/cache_manager/v1/test_cache_manager.py -v
python -m pytest tests/cache_manager/v1/test_transfer_manager.py -v
python -m pytest tests/engine/test_request.py -v
```

* [BugFix][KVCache] Fix BlockPool.allocate returns all blocks when num_blocks=0

## Motivation

当 `allocate(num_blocks=0)` 被调用时，Python 负索引陷阱导致严重错误：
`-0 == 0`，所以 `self._free_blocks[-0:]` 等价于 `self._free_blocks[0:]`，
会返回并清空整个空闲块列表，而非返回空列表。

## Modifications

在 `BlockPool.allocate` 中增加对 `num_blocks == 0` 的提前判断，直接返回 `[]`，
避免触发 Python 负索引陷阱。

## Usage or Command

```bash
# 运行相关单元测试验证修复
python -m pytest tests/cache_manager/v1/test_cache_manager.py -vv -s
```

* [KVCache][Test] add unit tests for cache_manager v1 modules

## Motivation

补全 cache_manager/v1 各模块的单测覆盖，确保核心方法有完整的测试保障。

## Modifications

新增/补充以下测试文件，全部 326 个用例通过：

- tests/cache_manager/v1/test_block_pool.py（新建）
  覆盖 BlockPool.get_metadata/set_metadata/resize、DeviceBlockPool/HostBlockPool
- tests/cache_manager/v1/test_metadata.py（新建）
  覆盖 BlockNode、RadixTreeStats、MatchResult、CacheSwapMetadata、AsyncTaskHandler
- tests/cache_manager/v1/test_cache_utils.py（补充）
  新增 hash_block_tokens、get_request_block_hasher、LayerDoneCounter 时间追踪及内部辅助方法
- tests/cache_manager/v1/test_radix_tree.py（补充）
  新增 TestCompleteSwapToDevice 专项测试类（6 个用例）
- tests/cache_manager/v1/test_cache_manager.py（补充）
  新增 offload_to_host、load_from_host、pending backup 系列、prepare_prefetch_metadata
- tests/cache_manager/v1/test_transfer_manager.py（补充）
  新增 _swap_single_layer 校验路径、sync_input/output_stream、record_input_stream_event

## Usage or Command

```bash
# 运行所有新增单测
source .venv/py310/bin/activate
python -m pytest tests/cache_manager/v1/test_block_pool.py \
  tests/cache_manager/v1/test_metadata.py \
  tests/cache_manager/v1/test_cache_utils.py \
  tests/cache_manager/v1/test_radix_tree.py \
  tests/cache_manager/v1/test_cache_manager.py \
  tests/cache_manager/v1/test_transfer_manager.py -v
# 期望结果：326 passed
```

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Co-authored-by: Jiang-Jia-Jun <163579578+Jiang-Jia-Jun@users.noreply.github.com>
 2026-04-21 14:39:00 +08:00
jc 6847891241 Mooncake storage register local buffer by chunk (#7416) 2026-04-17 10:39:34 +08:00
Jiang-Jia-Jun 26d6a20c2f [Optim] Remove IPCLock between CacheManager and WorkerProcess (#7299) 
* [Optim] Remove IPCLock between CacheManager and WorkerProcess

* Update envs.py

* Update worker_process.py

---------

Co-authored-by: jiang-jia-jun <jiangjiajun@baidu.com>
 2026-04-12 13:59:34 +08:00
jc 1cc0cf23c2 [BugFix] Set MC_MAX_MR_SIZE to avoid register hang in default (#7161) 
* Set MC_MAX_MR_SIZE to avoid register hang

* Set MC_MAX_MR_SIZE to avoid register hang
 2026-04-03 10:51:15 +08:00
Yonghua Li 98f3fc9267 [RL] [KVCache] let cache transfer managers update key prefix after weight update and add unit tests (#7083) 
* [test] add a few unit tests

* [feat] update key prefix when model weights are updated

* [test] try to fix test_worker_process
 2026-04-02 19:58:41 +08:00
jc af51fc46d6 [PD Disaggregation] Write the cache of preempted req to storage and refine PD Disaggregation (#7107) 
* Write the cache of preempted req to storage

* up

* fix
 2026-04-01 13:15:52 +08:00
kevin 18062c55bb [BugFix][KVCache] Fix mm hash boundary comparison in get_block_hash_extra_keys (#6929) 
* [BugFix][KVCache] Fix mm hash boundary comparison in get_block_hash_extra_keys

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* [BugFix][KVCache] Fix test_get_block_hash_extra_keys_boundary_cases assertions

## Motivation

测试用例 `test_get_block_hash_extra_keys_boundary_cases` 中，Block [4,8) 的
调用错误地传入了 `mm_idx=1`，跳过了 img0[2,5)；但 img0 覆盖 token 4，token 4
属于 block [4,8)，应被包含在 hash_keys 中。此外，所有 assertEqual 只校验了
hash_keys，未校验返回的 mm_idx 游标。

## Modifications

- `test_get_block_hash_extra_keys_boundary_cases`：
  - 改为链式调用，用上一次返回的 mm_idx 作为下一次入参，模拟真实调用循环
  - Block [4,8) 入参从 `mm_idx=1` 改为沿用上次返回的 `mm_idx=0`，期望值从 `[]` 改为 `["hash-0"]`
  - 所有断言改为 `assertEqual((mm_idx, hash_keys), (...))` 同时校验游标
- `test_get_block_hash_extra_keys_no_overlap_at_boundaries`：
  - Case B 入参从 `mm_idx=1` 改为 `mm_idx=0`（从头遍历，img-a 走 continue）
  - 所有断言增加 mm_idx 校验
- `test_get_block_hash_extra_keys_image_crosses_block_boundary`：
  - 所有断言增加 mm_idx 校验
- `test_get_block_hash_extra_keys_no_mm_inputs`：
  - 断言增加 mm_idx 校验
- `test_get_block_hash_extra_keys_handles_multimodal_segments`：
  - call2、call3 断言增加 mm_idx 校验

## Usage or Command

```bash
python -m pytest tests/cache_manager/test_prefix_cache_manager.py::TestPrefixCacheManagerCoverage -v -k "get_block_hash_extra_keys"
```

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

---------

Co-authored-by: chengyanfu <chengyanfu@baidu.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
 2026-03-30 17:13:31 +08:00
Jiang-Jia-Jun 1670b011a5 Revert "[BugFix] Add lock to avoid generating nan when using storage cache (#…" (#7075) 
This reverts commit 6d2ab8f2c0.
 2026-03-30 14:52:05 +08:00
jc 6d2ab8f2c0 [BugFix] Add lock to avoid generating nan when using storage cache (#7046) 
* Add lock to avoid generating nan

* up
 2026-03-30 14:50:32 +08:00
Dangweichong 3c9fd818e3 [BugFix] Fix RDMA initializes failed (#7025) 2026-03-26 17:45:39 +08:00
Yonghua Li a7f52c300d [Feature] support v1 update/clear api for RL (#6761) 
* [Feature] support v1 update/clear api for RL

* [fix] fix execute_model and add sleep/wakeup api

* [fix] fix mtp and key_prefix

* [chore] move _update_key_prefix to resume method

* [fix] make the interface safe to call multiple times

* [fix] fix some tiny bugs

* [chore] make small changes against pr review

* [docs] add docs for weight update

* [test] add some tests and update docs

* [style] fix code style check

* [test] fix ci

* [fix] fix stale control responses when control method timed out

* [chore] remove unused code

* [chore] fix code style

* [chore] optimize tags and key_prefix

* [test] fix ci

* [chore] fix code style

* [test] fix ci

* [fix] fix ep control

* [fix] fix ep control for engine cache queue
 2026-03-25 19:18:46 +08:00
jc bb881c2c0a [PD Disaggregation] pd + cache_storage support vl model (#6906) 
* pd + cache_storage support vl model

* support vl model

* fix test
 2026-03-23 15:35:20 +08:00
jc 950366e58d [PD Disaggregation][RL] Register to router with version and support rdma eager connect for pd (#6718) 
* [Feature] Register to router with version info for PD disaggregation

Add RegisterManager for PD (Prefill-Decode) disaggregated deployment:
- All instances (Prefill/Decode) register to Router with heartbeat
- Prefill instances fetch Decode instance list from Router
- Prefill instances establish eager RDMA connections to Decode instances
- Register info includes: host_ip, port, role, version, is_paused, connected_decodes

Changes:
- Add RegisterManager class for managing PD registration and RDMA connections
- Add version field to ModelConfig for model version tracking
- Add connected_decodes to register_info for tracking connected Decode instances
- Add FD_ENABLE_PD_RDMA_EAGER_CONNECT environment variable

Test fixes:
- Add None checks for load_config in FDConfig.__init__
- Add version attribute to test mock model configs

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* refine

* remove test

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
 2026-03-17 14:43:35 +08:00
gongweibao a6351dea0b [BugFix][Optimization] Replace silent failures with catchable exceptions and informative error messages (#6533) 
* init

* init

* fix format

* add

* add files

* add ut

* fix some

* add ut

* add more

* add

* fix pre-commit

* fix pre-commit

* fix cover

* skip long seq

* add

* add

* fix

* remove not need

* fix set attr

* fix comments

* fix comments

* fix failed tests

---------

Co-authored-by: gongweibao <gognweibao@baidu.com>
 2026-03-16 21:32:43 +08:00
jc 04fde3b227 [PD Disaggregation] Prefill and decode support cache storage (#6768) 
* Prefill and decode support cache storage

* up

* up

* update docs and refine mooncake store

* up
 2026-03-16 14:44:49 +08:00
jc 0466c7e8a8 Set MC_TCP_BIND_ADDRESS for mooncake store (#6782) 2026-03-11 16:56:39 +08:00
Jiang-Jia-Jun b05a6c4206 [BugFix][KVCache] Add inter-process lock to fix NaN error under DP+EP (#6724) 
* [BugFix] Support  to fix NaN bug in EP

* Optimze notion for all the funs

* Fix potential lock contention failure issues

* Update fastdeploy/inter_communicator/ipc_signal.py

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

* Update envs.py

* Update default value for USE_KVCACHE_LOCK

Change default value of USE_KVCACHE_LOCK from 1 to 0.

* Update worker_process.py

* Fix suffix wrong

* Update test_prefix_cache_manager.py

---------

Co-authored-by: Jiang-Jia-Jun <jiangjiajun@baidu.com>
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
 2026-03-10 21:55:32 +08:00
1 3a85ecf3bc [Others] Fix typos in log messages and comments (#6707) 
Fix spelling errors in log messages, docstrings, and comments:
- 'occured' -> 'occurred' (8 instances)
- 'Recieve'/'recieved' -> 'Receive'/'received' (7 instances)
- 'happend' -> 'happened' (3 instances)
- 'expet_servic' -> 'expert_service' (2 instances)
- 'meas' -> 'means' (1 instance)

No functional changes. Only log strings, docstrings, and comments are affected.

Co-authored-by: cloudforge1 <cloudforge1@users.noreply.github.com>
 2026-03-09 10:26:25 +08:00
jc b0fd242add [BugFix] Fix error in dynamic c8 cache (#6544) 
* [BugFix] Fix error in dynamic c8 cache

* fix device id
 2026-03-06 10:11:23 +08:00
Yonghua Li 27ae02fd82 [BugFix] fix prefix tree updating timeout (#6615) 2026-03-03 14:32:15 +08:00
RichardWooSJTU fe0b3a90ee [PD Disaggregation] Fix cache messager performance problem & add kv transfer benchmark tool (#6434) 
* fix cache messager performance problem

* dispatch param type
 2026-03-02 14:28:14 +08:00
kevin ecfd088a03 [BugFix] Add safety checks in recycle_gpu_blocks to prevent block allocation errors (#6531) 
* [BugFix] Add safety checks in recycle_gpu_blocks to prevent block allocation errors

- Check prefix tree status before recycling GPU blocks
- Validate gpu_block_ids is a list
- Add overflow check to prevent free block count exceeding total blocks

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* [BugFix] Fix AttributeError in recycle_gpu_blocks when prefix_tree_status_signal not initialized

- Add hasattr check before accessing prefix_tree_status_signal
- The signal is only initialized in launch_cache_messager, not in __init__
- Fixes CI test failure in test_prefix_cache_manager.py

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* [BugFix] Reset prefix cache when model weights are updating

- Call self.reset() before setting status to NORMAL in UPDATING state
- Ensure cache consistency when model weights change
- Consistent with CLEARING state handling

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
 2026-03-02 13:12:29 +08:00
Yonghua Li 7cf5e64c7a [BugFix] fix cache transfer manager init failed when using block_wise_fp8 and no storage backend (#6516) 
* [fix] fix cache transfer manager init failed when using block_wise_fp8 and no storage backend

* [fix] fix test_cache_transfer_manager

* [fix] fix test_cache_transfer_manager again

---------

Co-authored-by: Jiang-Jia-Jun <163579578+Jiang-Jia-Jun@users.noreply.github.com>
 2026-03-01 13:43:31 +08:00
jc 7b1d787b4b [BugFix] Fix storage_backend_type comparison bug in cache_transfer_manager.py (#6514) 
Co-authored-by: root <root@tjzj-inf-sci-k8s-hzz1-h62ni7-2178.tjzj.baidu.com>
 2026-02-26 19:32:24 +08:00
Yonghua Li e2332a1112 [BugFix] fix num_cpu_blocks computation (#6438) 
* [BugFix] fix num_cpu_blocks computation

* [fix] fix syntax and log

* [fix] pre-commit

* [fix] use getattr

* [fix] ci test
 2026-02-13 11:05:14 +08:00
CSWYF3634076 ec128068b7 [Others] Exit to ensure no residual processes (cpu cache & dp) (#6377) 
* [Others] good exit single dp

* [Others] good exit cpu cache dp>1

* [Others] good exit cpu cache dp>1 unittest
 2026-02-09 20:38:38 +08:00
Jiang-Jia-Jun 18e79dd660 [Metrics] Support cpu-cache-block-num (#6390) 
Co-authored-by: root <root@szzj-bcc-offline-1487319.szzj.baidu.com>
 2026-02-09 10:27:56 +08:00
Yonghua Li 5ac5ecd0b0 [BugFix] fix cache transfer tasks failure after cache cleared (#6202) 
* [fix] fix cache transfer tasks failure after cache cleared

* [fix] fix submit_task

* [fix] fix cache manager hang when clearing prefix cache

* [fix] fix list_proxy has no clear method

* [fix] fix barrier

* [fix] add barrier0

* [fix] add cache_task_is_paused_signal

* [fix] fix condition

* [fix] fix cache transfer  sync and delay prefix cache tree clearing

* [fix] fix typo

* [chore] polish code

* [fix] revert only rank0 write kv_cache_status_signal

* [fix] fix thread pool and prefix cache manager hang

* [fix] add timeout for task_swapping_event

* [fix] tolerate prefix cache manager error while prefix tree is cleared

* [chore] add more log

* [fix] fix test_prefix_cache_manager

* [fix] fix prefix_cache_status_signal usage
 2026-02-08 15:33:56 +08:00
jc d6b3c722c1 [KVCache] Storage cache supports c8 model (#6298) 
* Refine cache transfer manager
* Storage cache supports c8 model
 2026-02-06 12:01:17 +08:00
Moonchild1227 39dc4b0c2e [Feature] [KVCache] support file_store kv cache backend (#6188) 
* fix(examples): comment out stop.sh to avoid error when script is missing

* feat: add file_store support for cache manager

* [fix] fix multi gpu transfer

* [fix] fix global kvcache transfer

* [Feature] [KVCache] support file_store kv cache backend

* chore: update FileStore according to PR comments

* fix: remove comments

* fix: add swap_cache_layout for file store

* fix: remove rank key

* fix: Switch KV cache storage to pure file mode

* Temporarily disable support for Tensor types

* fix: remove args --kvcache_file_path & add envs FILE_BACKEND_STORAGE_DIR

* fixx: Simplify cache_transfer_manager.py

* fix: fix syntax bug

* fix: Simplify file_store.py

* fix: Use the key directly as the filename

* fix: Simplify set()

* fix: Simplify cache_transfer_manager.py & file_store.py

* fix: Only support load to cpu buffer

* feat: add FileStore backend for cache transfer

* fix: guard zmq import
 2026-02-03 14:37:58 +08:00
chenjian af1b1d2d56 [Feature] Support report token index by attention store (#6285) 
* [Feature] Support report token index by attention store

* fix format
 2026-02-02 10:41:11 +08:00
chenjian 292bab7e6d [BugFix] Fix bug for enable output caching (#6226) 
* [BugFix] Fix bug for enable output caching

* fix

* Fix

* fix

* fix ci
 2026-01-30 10:55:36 +08:00
jc b1698a79cb [RL] add version to the key of cache storage && refine raising error (#6160) 
* Waiting for cache transfer manager inited

* up

* up

* up

* up

* up

* fix according comments

* fix unittest

* fix

* fix unittest

* fix error

* pass storage_backend to worker
 2026-01-27 10:47:46 +08:00
Yonghua Li 833d00e2d7 [BugFix] move cache creation back to cache transfer process and adapt clear/update (#6144) 
* [fix] move cache creation back to cache transfer process

* [fix] fix clear cache

* [chore] change some log level

* [fix] fix clear cache

* [fix] fix clear cache for blockwisefp8 and mtp

* [fix] fix c8

* [fix] fix clear_mtp_cache args

* [chore] update cache_transfer_manager

* [fix] fix update mtp cache
 2026-01-24 21:59:13 +08:00
Yonghua Li 8d27a523e7 [Feature] [KVCache] support attention_store kv cache backend (#5823) 
* [feat] support attention_store kv cache backend

* [fix] fix codestyle

* [chore] optimize log

* [fix] fix write storage task

* [fix] fix read storage

* [fix] fix code conflict after merge develop

* [fix] fix cache bytes and read task token ids

* [chore] add model for cache transfer manager

* [chore] add some log

* [chore] remove launched_cache_manager_signal

* [fix] fix write_back_storage_task match_block_num condition

* [fix] fix swap_cost_time

* [ci] fix ci

* Update fastdeploy/engine/sched/resource_manager_v1.py

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

* Update fastdeploy/cache_manager/cache_transfer_manager.py

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

* Update fastdeploy/cache_manager/transfer_factory/mooncake_store/attention_store.py

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

---------

Co-authored-by: Jiang-Jia-Jun <163579578+Jiang-Jia-Jun@users.noreply.github.com>
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
 2026-01-22 21:01:23 +08:00
qwes5s5 b2a2e11551 [Feature] Support stopping the inference for the corresponding request in the online service after a disconnection request. (#5320) 
* request disconnect

* request disconnect

* fix bug

* fix bug--amend

---------

Co-authored-by: root <root@yq01-sys-rpm26xc1knu.yq01.baidu.com>
 2026-01-16 11:46:13 +08:00
Daci e10b51b8c6 [Feature] get_output_kv_signal blocking read mode & send_first_token (#5836) 
* get_output_kv_signal blocking read mode

* send first token before recycle

* xpu get_output_kv_signal blocking read mode

---------

Co-authored-by: Jiang-Jia-Jun <163579578+Jiang-Jia-Jun@users.noreply.github.com>
 2026-01-15 14:11:03 +08:00
Yonghua Li 456637002d [BugFix] fix cache transfer manager updating/clearing (#5930) 
* [fix] fix cache transfer manager updating/clearing

* [fix] fix code style

* [fix] fix config

* [fix] fix engine client

* [fix] let worker update kv cache status signal

* [fix] update worker process

* [fix] fix clear/update for case if comm group is shutdown

* [fix] update dynamic weight manager

* [fix] fix port

* [fix] add num_cpu_blocks arg for async_llm, and remove unnecessary waiting
 2026-01-13 05:09:29 -08:00
Yonghua Li 60ee72f682 [BugFix] [MultiAPIServer] fix rdma script and port check for multi api server (#5935) 
* [fix] fix rdma script and add more error log for multi api server

* [fix] log

* [fix] fix test_multi_api_server

* [fix] fix multi api server port check

---------

Co-authored-by: Jiang-Jia-Jun <163579578+Jiang-Jia-Jun@users.noreply.github.com>
 2026-01-12 10:38:52 +08:00
kevin 2d2b156252 [BugFix] fix dyc8 cache bug (#5958) 
* fix dyc8 cache bug

* update code
 2026-01-08 19:25:47 -08:00
kevin eabd01cd21 [BugFix] fix eb5 prefix bug (#5879) 
* fix eb5 prefix bug

* update ci test

* update code

* update code

* update code

* update code

* update code

* update code

* update code
 2026-01-06 23:50:39 -08:00
kevin a76e8ae40c [Feature] support rdma pd dy-c8 (#5788) 
* add rdma pd dy-c8

* update code
 2026-01-07 14:55:25 +08:00
Yonghua Li 9445fbe054 [KVCache] launch cache transfer processes only if hierarchical cache or kv cache storage is enabled (#5871) 
* [fix] temporarily forbid cpu cache in update/clear api

* [fix] stop launching cache transfer manager unless hierarchical cache is enabled

* [fix] fix no attr hierarchical cache

* [fix] fix ci

* [fix] fix test_prefix_cache_manager.py
 2026-01-06 14:27:47 +08:00
jc e9b25aa72f [BugFix] Storage backend gets env params (#5892) 
* Storage backend gets env params

* up

* up

* up
 2026-01-06 14:14:17 +08:00
jc e911ac2ce7 [BugFix] Refine the preparation of cpu and storage cache (#5777) 
* Refine the preparation of cpu and storage cache

* fix error

* fix error

* up

* fix

* up docs

* fix unittest

* remove debug info
 2026-01-05 10:13:30 +08:00
jc 95257c1dbd [Feature] RDMACommunicator send key and value scale (#5737) 
* RDMACommunicator send key and value scale
---------

Co-authored-by: kevin <chengyf112@gmail.com>
Co-authored-by: Jiang-Jia-Jun <163579578+Jiang-Jia-Jun@users.noreply.github.com>
 2026-01-05 10:04:24 +08:00
kevin 52dc9a7b85 [BugFix] skip mm revert (#5848) 
* skip mm revert

* update code

* update test
 2026-01-04 14:25:45 +08:00
MingkunZhang f732d7d2ad [Metax] adapt prefix caching & cpu swap (#5844) 
Co-authored-by: root <root@lt-wks-10-0-180-15.pub.metax-tech.com>
 2025-12-31 17:02:48 +08:00
周周周 7ae13b2326 [PD Disaggregation]remove unsed para in RDMACommManager (#5814) 2025-12-30 11:38:30 +08:00
kevin 5538dda3c8 [Feature] pd support dy-c8 ipc (#5750) 
* pd support dy-c8 ipc

* update code

* support v0

* update code
 2025-12-25 21:22:34 +08:00