perf(scale): 50k 压测 harness + resume 计数与 CI artifact 修复 - #3
Merged
Merged
Conversation
- perf_scale.rs: --start-doc 断点续跑参数（50k 压测期间遗留未提交） - perf_scale.rs: --max-contention-factor 断言，超标非零退出（供 CI 门槛，目标小于 2x） - 新增 .github/workflows/perf-scale.yml：手动触发 + 每周日 02:34 UTC 全量 50k，报告 JSON 上传 artifact - README/PERF_SCALE_50K.md：同步 50k 验证结论与 CI 复跑说明
resume 场景下 --start-doc 可能早于实际进度，重复写入的文档会替换旧 chunk，原实现用'库里已有数 + 本次写入数'累加会重复计数、虚报 indexed_chunks 与 chunks/s；改为以 DB 实际 chunk 数为准。 CI 侧给 artifact 上传加 if: always()：断言失败时才是最需要报告的时候，原来恰好在那时丢掉报告。
当前 workspace/UI/Tauri 版本均为 1.5.2，但 docs/release 只有到 v1.5.0，lint job 的版本一致性检查因此必然失败（上游 collab 本身也缺）。补齐 RELEASE_NOTES_v1.5.2.md，覆盖本次工程硬化与 OCR 变更、已知边界与升级说明。 同时把 docs/README.md 的 Release 索引补上 v1.5.0 与 v1.5.2 两条（此前 v1.5.0 也未登记）。
审查者指南本 PR 建立了可手动或定时复跑的 50k 压测 CI，并在 harness 中加入断点续跑、真实 chunk 计数和争用系数断言；同时修复失败时报告 artifact 丢失问题，补充规模验证文档及 1.5.2 release notes 以完成文档和版本一致性。 可恢复性能规模索引与报告的时序图sequenceDiagram
participant CI as PerfScaleCI
participant Harness as perf_scale
participant Store as VectorStore
participant DB as SQLiteDB
CI->>Harness: Run with --start-doc and --report
alt start_doc == 0
Harness->>Store: purge_all_index_data()
Harness->>Store: begin_full_rebuild()
else start_doc > 0
Harness->>Harness: seed_corpus(start_doc)
end
loop documents from start_doc to docs
Harness->>Store: index document chunks
Store->>DB: replace existing chunks
end
Harness->>Store: finish_full_rebuild()
Harness->>Store: load_from_db()
Store->>DB: count_chunks()
DB-->>Harness: actual chunk count
Harness->>Harness: write report with indexed_chunks
Harness-->>CI: contention assertion result
性能规模 CI 断言与 artifact 保留流程图flowchart TD
A[workflow_dispatch or weekly schedule] --> B[Run perf_scale with 50k defaults]
B --> C{contention_factor <= max limit?}
C -->|Yes| D[Process succeeds]
C -->|No| E[Process exits nonzero]
D --> F[Upload perf_scale_ci_report.json]
E --> F
F --> G[Artifact available for PASS or FAIL run]
文件级变更
提示与命令与 Sourcery 交互
自定义使用体验访问你的控制面板：
获取帮助Original review guide in EnglishReviewer's Guide本 PR 建立了可手动或定时复跑的 50k 压测 CI，并在 harness 中加入断点续跑、真实 chunk 计数和争用系数断言；同时修复失败时报告 artifact 丢失问题，补充规模验证文档及 1.5.2 release notes 以完成文档和版本一致性。 Sequence diagram for resumable perf-scale indexing and reportingsequenceDiagram
participant CI as PerfScaleCI
participant Harness as perf_scale
participant Store as VectorStore
participant DB as SQLiteDB
CI->>Harness: Run with --start-doc and --report
alt start_doc == 0
Harness->>Store: purge_all_index_data()
Harness->>Store: begin_full_rebuild()
else start_doc > 0
Harness->>Harness: seed_corpus(start_doc)
end
loop documents from start_doc to docs
Harness->>Store: index document chunks
Store->>DB: replace existing chunks
end
Harness->>Store: finish_full_rebuild()
Harness->>Store: load_from_db()
Store->>DB: count_chunks()
DB-->>Harness: actual chunk count
Harness->>Harness: write report with indexed_chunks
Harness-->>CI: contention assertion result
Flow diagram for perf-scale CI assertions and artifact preservationflowchart TD
A[workflow_dispatch or weekly schedule] --> B[Run perf_scale with 50k defaults]
B --> C{contention_factor <= max limit?}
C -->|Yes| D[Process succeeds]
C -->|No| E[Process exits nonzero]
D --> F[Upload perf_scale_ci_report.json]
E --> F
F --> G[Artifact available for PASS or FAIL run]
File-Level Changes
Tips and commandsInteracting with Sourcery
Customizing Your ExperienceAccess your dashboard to:
Getting Help
|
![sourcery-ai[bot]](https://avatars.githubusercontent.com/in/48477?s=60&v=4){kind=small}
There was a problem hiding this comment.
您好——我发现了 1 个问题
面向 AI Agent 的提示
请处理此次代码审查中的评论：
## 各条评论
### 评论 1
<location path="memori-core/examples/perf_scale.rs" line_range="195-197" />
<code_context>
- // 干净起点：删除旧 DB。
- let _ = std::fs::remove_file(&args.db_path);
+ // 干净起点：删除旧 DB（--start-doc>0 续跑时保留）。
+ if args.start_doc == 0 {
+ let _ = std::fs::remove_file(&args.db_path);
+ }
if let Some(parent) = args.db_path.parent() {
std::fs::create_dir_all(parent)?;
</code_context>
<issue_to_address>
**问题 (bug_risk)：** 当 `--start-doc` 大于零，但指定的数据库不存在时，代码会创建一个新的空数据库，并且只为 `start_doc..docs` 范围内的文档建立数据；此时，报告却声称包含完整的 `docs` 文档集，尽管前面的文档并不存在。
**触发条件：** 使用缺失、已删除或指定错误的 `--db-path` 开始续跑时。
**建议修复：** 除非数据库存在且包含预期的既有文档集，否则应立即失败；或者在继续之前，显式地为缺失的前缀文档建立数据。
</issue_to_address>Sourcery 评估
等待批准。 需要先处理 1 个发现的问题。
阻塞性发现：memori-core/examples/perf_scale.rs:197
Original comment in English
Hey - I've found 1 issue
Prompt for AI Agents
Please address the comments from this code review:
## Individual Comments
### Comment 1
<location path="memori-core/examples/perf_scale.rs" line_range="195-197" />
<code_context>
- // 干净起点：删除旧 DB。
- let _ = std::fs::remove_file(&args.db_path);
+ // 干净起点：删除旧 DB（--start-doc>0 续跑时保留）。
+ if args.start_doc == 0 {
+ let _ = std::fs::remove_file(&args.db_path);
+ }
if let Some(parent) = args.db_path.parent() {
std::fs::create_dir_all(parent)?;
</code_context>
<issue_to_address>
**issue (bug_risk):** When `--start-doc` is greater than zero but the specified database does not already exist, the code creates a new empty database and seeds only documents in `start_doc..docs`; the report then claims the full `docs` corpus even though the earlier documents are absent.
**Triggers:** When a resume is started with a missing, deleted, or incorrectly specified `--db-path`.
**Suggested fix:** Fail fast unless the database exists and contains the expected prior corpus, or explicitly seed the missing prefix before continuing.
</issue_to_address>Sourcery assessment
Approval pending. 1 finding to address first.
Blocking findings: memori-core/examples/perf_scale.rs:197
Comment on lines
+195
to
+197
| if args.start_doc == 0 { | ||
| let _ = std::fs::remove_file(&args.db_path); | ||
| } |
There was a problem hiding this comment.
问题 (bug_risk)： 当 --start-doc 大于零，但指定的数据库不存在时，代码会创建一个新的空数据库，并且只为 start_doc..docs 范围内的文档建立数据；此时，报告却声称包含完整的 docs 文档集，尽管前面的文档并不存在。
触发条件： 使用缺失、已删除或指定错误的 --db-path 开始续跑时。
建议修复： 除非数据库存在且包含预期的既有文档集，否则应立即失败；或者在继续之前，显式地为缺失的前缀文档建立数据。
Original comment in English
issue (bug_risk): When --start-doc is greater than zero but the specified database does not already exist, the code creates a new empty database and seeds only documents in start_doc..docs; the report then claims the full docs corpus even though the earlier documents are absent.
Triggers: When a resume is started with a missing, deleted, or incorrectly specified --db-path.
Suggested fix: Fail fast unless the database exists and contains the expected prior corpus, or explicitly seed the missing prefix before continuing.
This was referenced Sep 15, 2026
Owner
|
审过了，可以合。上一轮指出的两条都修对了，也都核实过：
其余部分（workflow 的手动输入 + 周期触发、 唯一一件事：请把 这份 release notes 在 #2 / #3 / #4 里完全相同（md5 一致），且写了 "adds OCR ingestion for images and scanned documents"。#4 的 OCR 还在返工（漏了间接 摘掉之后就可以合了。 |
Owner
|
复审通过，可以合了。
前一轮的两条修复维持不变，再确认一次：
无新增意见。 |
FPSZ
added a commit
to wjs480/Memori-Vault
that referenced
this pull request
Sep 16, 2026
合并 FPSZ#2/FPSZ#3/FPSZ#5 后的 collab，三处冲突均为双方各自新增内容，保留双方： - README.md「已实现但仍在优化」：50k 压测条目 + OCR 条目 - docs/qa/RETRIEVAL_BASELINE_V2.md：作答层 LLM-judge 基线 + OCR 接入与边界 - memori-server/src/dto.rs：AppSettings 的 index_filter 与 ocr_tesseract_path 两个字段 同时启用 pdf_with_indirect_resources_is_extracted（原 #[ignore]）： 手工构造 lopdf::Stream 用结构体字面量会绕过 /Length 写入（lopdf object.rs:602 的 Stream::new 才会 dict.set("Length", ...)），存盘后流长度为 0、load 回来 content 为空，解码拿不到像素，测试因此假失败——与 get_pages() 无关（实测 get_pages()=1、get_page_resources 也正确返回 inherited）。改用 Stream::new 后 测试通过，可作为间接 /Resources + 间接 /XObject 的回归锁。 Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
背景
评审对压测部分只提了两条：
if: always()与 resume 计数。本 PR 把压测 harness 与这两处修复放在一起。包含内容（3 个提交）
perf(scale): 压测 harness 支持续跑与断言，新增 50k CI 复跑 workflow（原提交）perf(scale): 修复压测 harness 的 resume 重复计数与 CI artifact 上传--start-doc早于实际进度时，重复写入的文档会替换旧 chunk；原实现用"库里已有数 + 本次写入数"累加 → 重复计数、虚报indexed_chunks与chunks/s。现改为以 DB 实际 chunk 数为准。if: always()：断言失败时才是最需要看报告的时候，原来恰好在那时丢掉 artifact。docs(release): 补齐 1.5.2 release notes 与文档索引—— 让 lint 的版本一致性检查能过（原因同 PR ①）验证
cargo test --workspace --all-targets全绿，0 失败collab最新050c690，behind_by=0Sourcery 总结
新增可恢复的 5 万规模性能验证，提供可靠的指标、即使失败也能保留的报告，以及更新后的发布文档。
新功能：
错误修复：
增强功能：
CI：
文档：
Original summary in English
Sourcery 总结
建立可恢复的 5 万规模性能验证机制，提供可靠的指标、在失败时保留报告的 CI，以及一致的发布文档。
新功能：
错误修复：
增强功能：
CI：
文档：
Original summary in English
Summary by Sourcery
Establish resumable 50k-scale performance validation with reliable metrics, failure-preserving CI reports, and aligned release documentation.
New Features:
Bug Fixes:
Enhancements:
CI:
Documentation: