Skip to content

feat(gateway): 图/视频调用经进程内 Gateway 做重试、Fallback 与观测 - #331

Open
xiaocheny214 wants to merge 17 commits into
1024XEngineer:mainfrom
xiaocheny214:feat/llm-gateway
Open

feat(gateway): 图/视频调用经进程内 Gateway 做重试、Fallback 与观测#331
xiaocheny214 wants to merge 17 commits into
1024XEngineer:mainfrom
xiaocheny214:feat/llm-gateway

Conversation

@xiaocheny214

@xiaocheny214 xiaocheny214 commented Aug 17, 2026 •
edited
Loading

Copy link
Copy Markdown
Contributor

Summary

  • 把图生图和 i2v 的重试、同家族 Fallback、熔断和结构化 trace 从 sufy.py 抽到 windup_framework.gateway。业务仍只调 ImageProvider.gen_image / VideoProvider.i2v。
  • 521/522/523/525 同型号只再打一次、不换型号；520/524 不重试。视频有 job_id 后不再 POST 第二单，仅 failed/cancelled 允许换型号新开。
  • 任务失败原因带上 request_id，便于从日志捞 attempt 链。积分扣费仍走现有 QUOTA_*。

Fixes #330

Test plan

  • backend/ 下 gateway + sufy + custom_action + executor 等相关测试 167 passed
  • 图像 522：同一型号恰好 2 次 POST，不打 fallback 型号
  • 图像 429：primary 最多 3 次后走 fallback
  • 图像 520：只打 1 次
  • 视频 submit 拿到 job_id 后，轮询 429/失败不再开第二单；上游 failed/cancelled 才换型号
  • 非法 video_model 仍 4xx
  • 失败任务的 error 含 request_id=

xiaocheny214 and others added 11 commits August 17, 2026 09:34
@xiaocheny214 @cursoragent
feat(gateway): 按状态码分类 UNREACHED/MAYBE_BILLED
2298818 
Co-authored-by: Cursor <[email protected]>
@xiaocheny214 @cursoragent
feat(gateway): 增加 fallback 与单价配置键
e25fbb8 
Co-authored-by: Cursor <[email protected]>
@xiaocheny214 @cursoragent
feat(gateway): 按 family 校验 scene 型号链
28fa46a 
Co-authored-by: Cursor <[email protected]>
@xiaocheny214 @cursoragent
feat(gateway): 522 只再试一次并按入口熔断
3a55aa1 
Co-authored-by: Cursor <[email protected]>
@xiaocheny214 @cursoragent
feat(gateway): 记录耗时与估算成本的 attempt trace
d5d5211 
Co-authored-by: Cursor <[email protected]>
@xiaocheny214 @cursoragent
refactor(providers): 图像 adapter 一次 POST 返回结构化结果
26992d7 
Co-authored-by: Cursor <[email protected]>
@xiaocheny214 @cursoragent
feat(gateway): 图像调用按策略重试与 Fallback
23fb730 
Co-authored-by: Cursor <[email protected]>
@xiaocheny214 @cursoragent
feat(gateway): 视频建单与跟单分离并按策略 Fallback
ef2363a 
Co-authored-by: Cursor <[email protected]>
@xiaocheny214 @cursoragent
fix(gateway): 已有 job_id 时禁止 429 再开第二单
7cef41f 
Co-authored-by: Cursor <[email protected]>
@xiaocheny214 @cursoragent
feat(gateway): 生成任务经 Gateway 装配并回写 request_id
4a9fe82 
Co-authored-by: Cursor <[email protected]>
@xiaocheny214 @cursoragent
fix(gateway): 熔断加锁并禁止无 job_id 时换型号
19fce6a 
并发过期不再 KeyError；跳过开路型号会记 fallback_used；视频分阶段耗时写入 trace；提交未拿到 job_id 时除 429 外不再换型号。

Co-authored-by: Cursor <[email protected]>
@vercel

vercel Bot commented Aug 17, 2026 •
edited
Loading

Copy link
Copy Markdown

The latest updates on your projects. Learn more about Vercel for GitHub.

1 Skipped Deployment
Project Deployment Actions Updated (UTC)
windup Ignored Ignored Preview Aug 19, 2026 9:19am

@codecov

codecov Bot commented Aug 17, 2026 •
edited
Loading

Copy link
Copy Markdown

Codecov Report

❌ Patch coverage is 88.95447% with 131 lines in your changes missing coverage. Please review.

Files with missing lines Patch % Lines
...ges/framework/src/windup_framework/gateway/chat.py 74.21% 49 Missing ⚠️
...es/framework/src/windup_framework/gateway/video.py 88.48% 22 Missing ⚠️
...es/framework/src/windup_framework/gateway/image.py 89.82% 17 Missing ⚠️
...s/framework/src/windup_framework/providers/sufy.py 82.82% 17 Missing ⚠️
...s/framework/src/windup_framework/gateway/ledger.py 75.00% 15 Missing ⚠️
...framework/src/windup_framework/gateway/registry.py 92.10% 3 Missing ⚠️
...es/framework/src/windup_framework/gateway/trace.py 97.11% 3 Missing ⚠️
...app/src/windup_app/server/orchestrator/executor.py 93.75% 2 Missing ⚠️
.../framework/src/windup_framework/config/provider.py 96.77% 1 Missing ⚠️
...framework/src/windup_framework/gateway/classify.py 97.05% 1 Missing ⚠️
... and 1 more

📢 Thoughts on this report? Let us know!

fennoai[bot]
fennoai Bot reviewed Aug 17, 2026
View reviewed changes

@fennoai fennoai Bot left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Review conclusion

整体分层与提交后不重复建单的主路径一致，但视频错误状态仍有两处会扩大重试或污染后续路由：轮询 52x 被 adapter 与 Gateway 双重重试，以及本次请求明确禁止 fallback 时仍提前打开型号熔断。建议修正这两个状态转换后再合入。

验证：完整检查了固定 diff 4103106...19fce6a 的 gateway、Sufy adapter、executor/API 接入与相关测试；git diff --check 通过。当前运行环境缺少 uv 和项目 Python 依赖，未能执行 pytest。

View job run

Comment thread backend/packages/framework/src/windup_framework/gateway/video.py Outdated
Comment thread backend/packages/framework/src/windup_framework/providers/sufy.py
@johnnyzhang-eng johnnyzhang-eng mentioned this pull request Aug 17, 2026
Open
@johnnyzhang-eng

johnnyzhang-eng commented Aug 17, 2026

Copy link
Copy Markdown
Contributor

我的 #324 也改 sufy.py——抽了个 ChatCompletionsFace 让出图和判官共用建连和重试，和你这版重叠。你把重试整个移到 Gateway 之后那层就多余了，所以不管哪边先合，我都按你的 AdapterResult 收，不另起一套。

@xiaocheny214

xiaocheny214 commented Aug 17, 2026

Copy link
Copy Markdown
Contributor Author

我的 #324 也改 sufy.py——抽了个 ChatCompletionsFace 让出图和判官共用建连和重试，和你这版重叠。你把重试整个移到 Gateway 之后那层就多余了，所以不管哪边先合，我都按你的 AdapterResult 收，不另起一套。

现按照我的来做吧。咱们把重试的逻辑统一放在gateway做，这样方便后续的链路追踪、熔断、限流等逻辑的引入，以及更换模型路由。这样看起来，重试逻辑也不会太过散乱，你觉得呢？
@johnnyzhang-eng

This was referenced Aug 17, 2026
Open
Open
Closed
@johnnyzhang-eng

johnnyzhang-eng commented Aug 18, 2026

Copy link
Copy Markdown
Contributor

同意，重试归 Gateway。#324 没有另起一套：它改的是 sufy.py 里已有那处，把措辞从「图像服务」改成通用、让判官复用同一条建连和退避，你这版落地后删的是同一处，不会留两套。另外 #331 自 08-17 是冲突态没再动过而 #324 已绿，所以我按互不阻塞处理，谁先合我跟谁改。

@xiaocheny214
xiaocheny214 requested a review from minorcell August 19, 2026 08:45
@xiaocheny214
feat(gateway): 同入口多 key 切换并拆分 attempt 冷热字段
c47a057 
429 换 key 不换模型，522 跳过该入口剩余 key；Chat 走同一套路由。
AttemptTrace 只保留热字段，排障详情进 AttemptDetail。
xyh202131
xyh202131 previously approved these changes Aug 19, 2026
View reviewed changes
xyh202131
xyh202131 previously approved these changes Aug 19, 2026
View reviewed changes

@xyh202131 xyh202131 left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

批准：重试、Fallback、熔断与 trace 集中到 Gateway，且视频拿到 job_id 后不重复下单，避免重复扣费；请求链可追踪，错误分类覆盖完整。

@xiaocheny214
fix(gateway): 去掉合入后重复的 generate 调用
6474cab 
路线选择前误用未赋值的 master，且不应再把型号传给 _get_generator。
xyh202131
xyh202131 previously approved these changes Aug 19, 2026
View reviewed changes

@xyh202131 xyh202131 left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

批准：重试、Fallback、熔断与 trace 集中到 Gateway，视频拿到 job_id 后不重复下单；最新修正也去除了重复 generate 调用，避免重复请求与扣费。

@xiaocheny214
fix(gateway): 对齐合入后的 chat 工厂与预付费结算
ce2de8c 
Chat 工厂校验 AI_API_KEY / AI_CHAT_MODEL；图片失败解冻积分；stance 测试改打 generate_character_action。
@vercel

vercel Bot commented Aug 19, 2026

Copy link
Copy Markdown

Deployment failed for project windup with the following error:

Resource is limited - try again in 24 hours (more than 100, code: "api-deployments-free-per-day").

Learn More: https://vercel.com/huyan-s-projects1?upgradeToPro=build-rate-limit

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@fennoai fennoai[bot] fennoai[bot] left review comments
@xyh202131 xyh202131 xyh202131 left review comments
@minorcell minorcell Awaiting requested review from minorcell

At least 1 approving review is required to merge this pull request.

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

图/视频调用需要把重试、Fallback 和计费边界从 adapter 里抽出来

3 participants

@xiaocheny214 @johnnyzhang-eng @xyh202131