fix(vision-v2): P0全部识别失败 - timeout过短+缺response_format #2208

Merged
xiaoxia merged 1 commits from fix/vision-v2-timeout into develop 2026-10-05 21:10:05 +08:00
Owner

P0根因

staging实测3图全部返回「未识别/无法判断」,根因:

  1. timeout设置太短(fast_json=8s/外层8s/pro=20s),qwen3.8-flash关thinking后单图实测8-9s,staging网络稍慢全部超时cancel
  2. 缺少response_format={"type":"json_object"},qwen偶发输出中文解释而非JSON
  3. 超时后走pro兜底,pro也可能因20s超时再次失败

修复

  • fast_json: _DEFAULT_TIMEOUT 8→12s
  • fast_path: _FAST_TIMEOUT/_FAST_JSON_TIMEOUT 8→12s, _PRO_TIMEOUT 20→25s
  • vlm_fallback: _DEFAULT_TIMEOUT 20→25s
  • fast_json + fallback 都加response_format={"type":"json_object"}强约束JSON输出(实测+2s加速,更稳定)

本地E2E验证(云电脑直连DashScope)

3图(unsplash人像):

  • 总耗时 8.57s,3/3 fast_json命中,无pro兜底
  • 单图elapsed: 5.73s / 4.46s / 8.56s
  • portrait_prompt均正常输出人物穿搭描述

立刻合入staging验证。

## P0根因 staging实测3图全部返回「未识别/无法判断」,根因: 1. timeout设置太短(fast_json=8s/外层8s/pro=20s),qwen3.8-flash关thinking后单图实测8-9s,staging网络稍慢全部超时cancel 2. 缺少response_format={"type":"json_object"},qwen偶发输出中文解释而非JSON 3. 超时后走pro兜底,pro也可能因20s超时再次失败 ## 修复 - fast_json: _DEFAULT_TIMEOUT 8→12s - fast_path: _FAST_TIMEOUT/_FAST_JSON_TIMEOUT 8→12s, _PRO_TIMEOUT 20→25s - vlm_fallback: _DEFAULT_TIMEOUT 20→25s - fast_json + fallback 都加response_format={"type":"json_object"}强约束JSON输出(实测+2s加速,更稳定) ## 本地E2E验证(云电脑直连DashScope) 3图(unsplash人像): - 总耗时 **8.57s**,3/3 fast_json命中,无pro兜底 - 单图elapsed: 5.73s / 4.46s / 8.56s - portrait_prompt均正常输出人物穿搭描述 立刻合入staging验证。
xiaoxia added 1 commit 2026-10-05 20:36:14 +08:00
fix(vision-v2): P0 全部识别失败 - timeout 过短+缺 response_format
CI/CD Pipeline / Check push changed paths (pull_request) Has been skipped
CI/CD Pipeline / Check if frontend-only change (pull_request) Successful in 2s
CI/CD Pipeline / Dedup Check - skip PR tests when covered by push pipeline (pull_request) Successful in 2s
CI/CD Pipeline / Frontend Lint (pull_request) Has been skipped
CI/CD Pipeline / Frontend Unit Tests (pull_request) Has been skipped
CI/CD Pipeline / PR Build Web Image (pull_request) Has been skipped
CI/CD Pipeline / Build Staging API Image (pull_request) Has been skipped
CI/CD Pipeline / Build Staging Web Image (pull_request) Has been skipped
CI/CD Pipeline / Build Staging Worker Image (pull_request) Has been skipped
CI/CD Pipeline / Retag skipped Staging API Image (pull_request) Has been skipped
CI/CD Pipeline / Retag skipped Staging Web Image (pull_request) Has been skipped
CI/CD Pipeline / Retag skipped Staging Worker Image (pull_request) Has been skipped
CI/CD Pipeline / Deploy Staging (Watchtower auto-deploy) (pull_request) Has been skipped
CI/CD Pipeline / Staging E2E Tests (pull_request) Has been skipped
CI/CD Pipeline / Staging API Integration Tests (pull_request) Has been skipped
CI/CD Pipeline / ACR Image Cleanup (pull_request) Has been skipped
CI/CD Pipeline / PR Build API Image (pull_request) Successful in 2m18s
Preview Deploy / Deploy Preview Environment (pull_request) Successful in 2m33s
PR Automation / Auto Approve on CI Green (pull_request) Successful in 3m9s
CI/CD Pipeline / PR Build Worker Image (pull_request) Successful in 3m38s
AI Code Review / AI Code Review (pull_request) Successful in 6m57s
CI/CD Pipeline / Integration Tests (pull_request) Successful in 9m24s
CI/CD Pipeline / Validate - Python (mypy + alembic) (pull_request) Successful in 12m43s
PR Automation / Auto Merge on CI Green + Approved (pull_request) Successful in 10m48s
CI/CD Pipeline / Validate - Style (pull_request) Failing after 15m24s
CI/CD Pipeline / Unit Tests (pull_request) Failing after 17m38s
CI/CD Pipeline / Validate - Security (pull_request) Successful in 26m35s
CI/CD Pipeline / Build Production API Image (pull_request) Has been skipped
CI/CD Pipeline / Build Production Web Image (pull_request) Has been skipped
CI/CD Pipeline / Build Production Worker Image (pull_request) Has been skipped
CI/CD Pipeline / Deploy Production (pull_request) Has been skipped
CI/CD Pipeline / Canary Release to Production (pull_request) Has been skipped
CI/CD Pipeline / CI Gate (pull_request) Failing after 2s
CI/CD Pipeline / Production Browser E2E (pull_request) Has been skipped
ACR Cleanup / ACR Image Cleanup (pull_request_target) Successful in 21s
Preview Cleanup / Cleanup Preview Environment (pull_request) Successful in 1m28s
65343473d8
根因:
- _FAST_TIMEOUT/_FAST_JSON_TIMEOUT 设为8s,但qwen3.8-flash关thinking后单图实测8-9s,staging网络稍慢即全部超时
- 超时cancel后走pro兜底,pro timeout=20s也偏紧
- 缺少 response_format=json_object 导致qwen偶发输出中文解释而非JSON

修复:
- fast_json: _DEFAULT_TIMEOUT 8→12s
- fast_path: _FAST_TIMEOUT/_FAST_JSON_TIMEOUT 8→12s, _PRO_TIMEOUT 20→25s
- vlm_fallback: _DEFAULT_TIMEOUT 20→25s
- fast_json + fallback 都加 response_format={'type':'json_object'}强约束JSON输出

本地3图E2E: 8.57s, 3/3 fast_json命中, portrait_prompt正常输出一位年轻男性/女性...

Refs: staging P0 3/3全返回未识别/无法判断

🚀 预览环境已部署

项目 详情
PR号 #2208
预览链接 https://pr-2208.preview.xiaoxiajianji.com
API环境 staging

💡 预览环境使用 staging API 数据,请勿在预览环境中操作重要数据。

🔄 每次提交新代码后预览环境会自动更新。

🗑️ PR 关闭或合并后,预览环境会自动清理。

🚀 **预览环境已部署** | 项目 | 详情 | |------|------| | PR号 | #2208 | | 预览链接 | [https://pr-2208.preview.xiaoxiajianji.com](https://pr-2208.preview.xiaoxiajianji.com) | | API环境 | staging | > 💡 预览环境使用 staging API 数据,请勿在预览环境中操作重要数据。 > > 🔄 每次提交新代码后预览环境会自动更新。 > > 🗑️ PR 关闭或合并后,预览环境会自动清理。
xiaoxia merged commit 7adbb7d331 into develop 2026-10-05 21:10:05 +08:00

🗑️ 预览环境已清理

PR #2208 已关闭或合并,对应的预览环境已被清理。

如有需要,可以重新打开 PR 来重新生成预览环境。

🗑️ **预览环境已清理** PR #2208 已关闭或合并,对应的预览环境已被清理。 > 如有需要,可以重新打开 PR 来重新生成预览环境。
Sign in to join this conversation.