feat(blog): DB 장애 원인과 인덱스 튜닝 글 추가 - #86
Conversation
슬로우 쿼리와 rows_examined, EXPLAIN을 바탕으로\n인덱스 설계 문제를 다룬 새 글을 추가했어요.\n\nAI 초안을 블로그 톤에 맞게 재구성하고\n과한 단정은 줄여 기술적 맥락을 보강했어요.
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: e5d5b08a2d
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| SUM_ROWS_EXAMINED AS rows_examined, | ||
| SUM_ROWS_SENT AS rows_sent |
There was a problem hiding this comment.
rows_examined here is a digest total
performance_schema.events_statements_summary_by_digest stores aggregated counters per normalized query, so SUM_ROWS_EXAMINED and SUM_ROWS_SENT are totals across all executions. When COUNT_STAR > 1, the later conclusion that a single lookup scanned “74만 건” is no longer supported by this query and can overstate the per-request cost by orders of magnitude. If you want to explain one request’s footprint, this needs an average such as SUM_ROWS_EXAMINED / COUNT_STAR or wording that these are cumulative totals.
Useful? React with 👍 / 👎.
| INDEX_NAME, | ||
| GROUP_CONCAT(COLUMN_NAME) AS indexed_columns | ||
| FROM INFORMATION_SCHEMA.STATISTICS | ||
| WHERE TABLE_SCHEMA = 'tableName' |
There was a problem hiding this comment.
STATISTICS lookup
In INFORMATION_SCHEMA.STATISTICS, TABLE_SCHEMA is the database/schema name, not the table name. As written, readers who follow the placeholder literally for the activity_queue example will get no rows back and won’t be able to reproduce the index-inspection step. This should use the schema name (and optionally add TABLE_NAME = 'activity_queue' if the intent is to inspect just that table).
Useful? React with 👍 / 👎.
1 participant
변경 내용
meta.json추가로 slug, description, tags 등록의도
영향 범위
검증
npm run buildnpm test체크리스트
codex/<task>)에서 진행함<type>: <한국어 설명>) 준수