add .gitignore, update README with new API server details, and add requirements.txt for dependencies
This commit is contained in:
iHeyTang
parent
6cc9aeab32
commit
ab12c1482d
|
|
@ -0,0 +1 @@
|
|||
.DS_Store
|
||||
39
README.md
39
README.md
|
|
@ -1,21 +1,24 @@
|
|||
# ComfyUI 工作流管理
|
||||
|
||||
|
||||
|
||||
## 功能
|
||||
|
||||
- 工作流上传
|
||||
- 工作流版本控制
|
||||
- 工作流加载
|
||||
- 需配置工作流服务器 (详见**WAAS(工作流即服务) Demo API服务器**)
|
||||
- 需配置工作流服务器 (详见**WAAS(工作流即服务) Demo API 服务器**)
|
||||
|
||||
# WAAS(工作流即服务) Demo API 服务器
|
||||
|
||||
# WAAS(工作流即服务) Demo API服务器
|
||||
|
||||
|
||||
- 路径: ./workflow_service
|
||||
- 部署: ```modal deploy modal_deploy.py --name waas-demo -e dev```
|
||||
- 配置: 参考```.env```文件
|
||||
- 部署: `modal deploy modal_deploy.py --name waas-demo -e dev`
|
||||
- 配置: 参考`.env`文件
|
||||
- 必须按照指定规则命名
|
||||
- **输入节点名**: 前缀 **INPUT_**
|
||||
- **除生成文件节点外输出节点名**: 前缀 **OUTPUT_**
|
||||
- **输入节点名**: 前缀 **INPUT\_**
|
||||
- **除生成文件节点外输出节点名**: 前缀 **OUTPUT\_**
|
||||
- 支持输入节点:
|
||||
- comfyui-core
|
||||
- 加载图像
|
||||
|
|
@ -26,7 +29,7 @@
|
|||
- 字符串
|
||||
- 浮点数
|
||||
- 支持输出节点:
|
||||
- 所有在output文件夹中生成文件(图片/视频)的节点
|
||||
- 所有在 output 文件夹中生成文件(图片/视频)的节点
|
||||
- comfyui-easy-use
|
||||
- 展示任何
|
||||
- 数据库
|
||||
|
|
@ -41,10 +44,12 @@
|
|||
)
|
||||
```
|
||||
- 路由
|
||||
|
||||
- GET /api/workflow: 列出工作流
|
||||
- POST /api/workflow: 添加工作流
|
||||
- DELETE /api/workflow: 删除工作流
|
||||
- GET /api/run/{base_name}: 获取工作流输入输出元数据
|
||||
|
||||
```
|
||||
输入:
|
||||
*base_name: 工作流名称
|
||||
|
|
@ -75,7 +80,9 @@
|
|||
}
|
||||
}
|
||||
```
|
||||
|
||||
- POST /api/run/{base_name}: 执行工作流
|
||||
|
||||
```
|
||||
输入:
|
||||
*base_name: 工作流名称
|
||||
|
|
@ -92,3 +99,21 @@
|
|||
]
|
||||
}
|
||||
```
|
||||
|
||||
# JS 脚本使用指南
|
||||
|
||||
### MacOS
|
||||
|
||||
将这个脚本直接放在该目录下，重启 ComfyUI 即可
|
||||
/Applications/ComfyUI.app/Contents/Resources/ComfyUI/custom_nodes/ComfyUI-Manager/js
|
||||
|
||||
### Windows
|
||||
|
||||
将这个脚本直接放在该目录下，重启 ComfyUI 即可
|
||||
{Comfy 的安装目录}\comfy\ComfyUI\custom_nodes\ComfyUI-Manager\js
|
||||
|
||||
### Q&A
|
||||
|
||||
主要是找到 `custom_nodes/ComfyUI-Manager` 的位置，这个目录下有一个 js 目录，把 `publisher.js` 脚本直接扔到这个 js 目录下即可。
|
||||
|
||||
一些版本安装的时候，用户额外装的的 custom_nodes 和系统自带的可能还不在一个位置，比如 MacOS 装客户端版本。用户额外装的是在 Documents 下，但 ComfyUI-Manager 是在对应的安装包目录里，得去 `/Applications` 里找。Windows 可能也会有类似的逻辑
|
||||
|
|
|
|||
|
|
@ -12,6 +12,7 @@ from workflow_service.config import Settings, ComfyUIServer
|
|||
|
||||
settings = Settings()
|
||||
|
||||
|
||||
# [新增] 定义一个自定义异常，用于封装来自ComfyUI的执行错误
|
||||
class ComfyUIExecutionError(Exception):
|
||||
def __init__(self, error_data: dict):
|
||||
|
|
@ -24,7 +25,10 @@ class ComfyUIExecutionError(Exception):
|
|||
)
|
||||
super().__init__(message)
|
||||
|
||||
async def get_server_status(server: ComfyUIServer, session: aiohttp.ClientSession) -> Dict[str, Any]:
|
||||
|
||||
async def get_server_status(
|
||||
server: ComfyUIServer, session: aiohttp.ClientSession
|
||||
) -> Dict[str, Any]:
|
||||
"""
|
||||
检查单个ComfyUI服务器的详细状态。
|
||||
返回一个包含可达性、队列状态和详细队列内容的字典。
|
||||
|
|
@ -33,10 +37,7 @@ async def get_server_status(server: ComfyUIServer, session: aiohttp.ClientSessio
|
|||
status_info = {
|
||||
"is_reachable": False,
|
||||
"is_free": False,
|
||||
"queue_details": {
|
||||
"running_count": 0,
|
||||
"pending_count": 0
|
||||
}
|
||||
"queue_details": {"running_count": 0, "pending_count": 0},
|
||||
}
|
||||
try:
|
||||
queue_url = f"{server.http_url}/queue"
|
||||
|
|
@ -51,10 +52,10 @@ async def get_server_status(server: ComfyUIServer, session: aiohttp.ClientSessio
|
|||
|
||||
status_info["queue_details"] = {
|
||||
"running_count": running_count,
|
||||
"pending_count": pending_count
|
||||
"pending_count": pending_count,
|
||||
}
|
||||
|
||||
status_info["is_free"] = (running_count == 0 and pending_count == 0)
|
||||
status_info["is_free"] = running_count == 0 and pending_count == 0
|
||||
|
||||
except Exception as e:
|
||||
# 当请求失败时，将返回上面定义的、结构正确的初始 status_info
|
||||
|
|
@ -82,21 +83,38 @@ async def select_server_for_execution() -> ComfyUIServer:
|
|||
|
||||
if free_servers:
|
||||
selected_server = random.choice(free_servers)
|
||||
print(f"发现 {len(free_servers)} 个空闲服务器。已选择: {selected_server.http_url}")
|
||||
print(
|
||||
f"发现 {len(free_servers)} 个空闲服务器。已选择: {selected_server.http_url}"
|
||||
)
|
||||
return selected_server
|
||||
else:
|
||||
# 后备方案：选择一个可达的服务器，即使它很忙
|
||||
reachable_servers = [servers[i] for i, status in enumerate(results) if status["is_reachable"]]
|
||||
reachable_servers = [
|
||||
servers[i] for i, status in enumerate(results) if status["is_reachable"]
|
||||
]
|
||||
if reachable_servers:
|
||||
selected_server = random.choice(reachable_servers)
|
||||
print(f"所有服务器当前都在忙。从可达服务器中随机选择: {selected_server.http_url}")
|
||||
print(
|
||||
f"所有服务器当前都在忙。从可达服务器中随机选择: {selected_server.http_url}"
|
||||
)
|
||||
return selected_server
|
||||
else:
|
||||
# 最坏情况：所有服务器都不可达，抛出异常
|
||||
raise ConnectionError("所有配置的ComfyUI服务器都不可达。")
|
||||
|
||||
|
||||
async def queue_prompt(prompt: dict, client_id: str, http_url: str) -> str:
|
||||
async def execute_prompt_on_server(prompt: Dict, server: ComfyUIServer) -> Dict:
|
||||
"""
|
||||
在指定的服务器上执行一个准备好的prompt。
|
||||
"""
|
||||
client_id = str(uuid.uuid4())
|
||||
prompt_id = await _queue_prompt(prompt, client_id, server.http_url)
|
||||
print(f"工作流已在 {server.http_url} 上入队，Prompt ID: {prompt_id}")
|
||||
results = await _get_execution_results(prompt_id, client_id, server.ws_url)
|
||||
return results
|
||||
|
||||
|
||||
async def _queue_prompt(prompt: dict, client_id: str, http_url: str) -> str:
|
||||
"""通过HTTP POST将工作流任务提交到指定的ComfyUI服务器。"""
|
||||
for node_id in prompt:
|
||||
prompt[node_id]["inputs"][f"cache_buster_{uuid.uuid4().hex}"] = random.random()
|
||||
|
|
@ -111,7 +129,7 @@ async def queue_prompt(prompt: dict, client_id: str, http_url: str) -> str:
|
|||
return result["prompt_id"]
|
||||
|
||||
|
||||
async def get_execution_results(prompt_id: str, client_id: str, ws_url: str) -> dict:
|
||||
async def _get_execution_results(prompt_id: str, client_id: str, ws_url: str) -> dict:
|
||||
"""
|
||||
通过WebSocket连接到指定的ComfyUI服务器，聚合执行结果。
|
||||
[核心改动] 新增对 'execution_error' 消息的处理。
|
||||
|
|
@ -128,26 +146,26 @@ async def get_execution_results(prompt_id: str, client_id: str, ws_url: str) ->
|
|||
continue
|
||||
|
||||
message = json.loads(out)
|
||||
msg_type = message.get('type')
|
||||
data = message.get('data')
|
||||
msg_type = message.get("type")
|
||||
data = message.get("data")
|
||||
|
||||
if not (data and data.get('prompt_id') == prompt_id):
|
||||
if not (data and data.get("prompt_id") == prompt_id):
|
||||
continue
|
||||
|
||||
# [核心改动] 捕获并处理执行错误
|
||||
if msg_type == 'execution_error':
|
||||
if msg_type == "execution_error":
|
||||
print(f"ComfyUI执行错误 (Prompt ID: {prompt_id}): {data}")
|
||||
# 抛出自定义异常，将错误详情传递出去
|
||||
raise ComfyUIExecutionError(data)
|
||||
|
||||
if msg_type == 'executed':
|
||||
node_id = data.get('node')
|
||||
output_data = data.get('output')
|
||||
if msg_type == "executed":
|
||||
node_id = data.get("node")
|
||||
output_data = data.get("output")
|
||||
if node_id and output_data:
|
||||
aggregated_outputs[node_id] = output_data
|
||||
print(f"收到节点 {node_id} 的输出 (Prompt ID: {prompt_id})")
|
||||
|
||||
elif msg_type == 'executing' and data.get('node') is None:
|
||||
elif msg_type == "executing" and data.get("node") is None:
|
||||
print(f"Prompt ID: {prompt_id} 执行完成。")
|
||||
return aggregated_outputs
|
||||
|
||||
|
|
@ -161,17 +179,9 @@ async def get_execution_results(prompt_id: str, client_id: str, ws_url: str) ->
|
|||
raise e
|
||||
|
||||
except websockets.exceptions.InvalidURI as e:
|
||||
print(f"错误: 尝试连接的WebSocket URI无效: '{full_ws_url}'. 原始URL: '{ws_url}'. 错误: {e}")
|
||||
print(
|
||||
f"错误: 尝试连接的WebSocket URI无效: '{full_ws_url}'. 原始URL: '{ws_url}'. 错误: {e}"
|
||||
)
|
||||
raise e
|
||||
|
||||
return aggregated_outputs
|
||||
|
||||
async def execute_prompt_on_server(prompt: Dict, server: ComfyUIServer) -> Dict:
|
||||
"""
|
||||
在指定的服务器上执行一个准备好的prompt。
|
||||
"""
|
||||
client_id = str(uuid.uuid4())
|
||||
prompt_id = await queue_prompt(prompt, client_id, server.http_url)
|
||||
print(f"工作流已在 {server.http_url} 上入队，Prompt ID: {prompt_id}")
|
||||
results = await get_execution_results(prompt_id, client_id, server.ws_url)
|
||||
return results
|
||||
Loading…
Reference in New Issue