{"id":108235,"date":"2026-09-21T11:24:42","date_gmt":"2026-09-21T03:24:42","guid":{"rendered":"https:\/\/www.wsisp.com\/helps\/108235.html"},"modified":"2026-09-21T11:24:42","modified_gmt":"2026-09-21T03:24:42","slug":"%e5%ae%9e%e6%88%98%ef%bd%9c10-%e4%b8%87%e8%a1%8c%e4%ba%91%e5%87%bd%e6%95%b0%e6%97%a5%e5%bf%97%e7%9a%84%e8%80%83%e5%8f%a4%e5%bc%8f%e6%8e%92%e9%9a%9c%ef%bc%9a%e7%94%a8%e8%85%be","status":"publish","type":"post","link":"https:\/\/www.wsisp.com\/helps\/108235.html","title":{"rendered":"\u5b9e\u6218\uff5c10 \u4e07\u884c\u4e91\u51fd\u6570\u65e5\u5fd7\u7684\u201c\u8003\u53e4\u5f0f\u201c\u6392\u969c\uff1a\u7528\u817e\u8baf\u4e91\u52a9\u624b\u628a\u5076\u53d1\u8d85\u65f6\u538b\u7f29\u6210\u4e00\u6761\u6839\u56e0\u94fe"},"content":{"rendered":"

\u5b9e\u6218｜10 \u4e07\u884c\u4e91\u51fd\u6570\u65e5\u5fd7\u7684"\u8003\u53e4\u5f0f"\u6392\u969c：\u7528\u817e\u8baf\u4e91\u52a9\u624b\u628a\u5076\u53d1\u8d85\u65f6\u538b\u7f29\u6210\u4e00\u6761\u6839\u56e0\u94fe<\/h2>\n

\u5206\u7c7b\u4e13\u680f：\u817e\u8baf\u4e91\u52a9\u624b\u5b9e\u6218 \/ \u53ef\u89c2\u6d4b\u6027
\n\u6807\u7b7e：\u817e\u8baf\u4e91\u52a9\u624b\u3001\u4e91\u51fd\u6570 SCF\u3001\u65e5\u5fd7\u670d\u52a1 CLS\u3001\u5076\u53d1\u6545\u969c\u3001P99\u3001\u6839\u56e0\u5206\u6790\u3001Python
\n\u6458\u8981：\u5076\u53d1\u8d85\u65f6（0.3% \u6982\u7387）\u662f\u6392\u969c\u91cc\u6700\u96be\u7684\u4e00\u7c7b：\u56fe\u8868\u770b\u5747\u503c\u770b\u4e0d\u51fa\u6765，\u62bd\u6837\u53c8\u62bd\u4e0d\u4e2d\u3002\u672c\u6587\u7ed9\u51fa\u300c\u65e5\u5fd7\u805a\u5408 \u2192 \u5c3e\u90e8\u6837\u672c\u753b\u50cf \u2192 \u7279\u5f81\u5bf9\u6bd4\u8868 \u2192 AI \u5f52\u56e0\u300d\u7684\u8003\u53e4\u5f0f\u6d41\u7a0b，\u628a 10 \u4e07\u884c\u65e5\u5fd7\u6536\u655b\u5230\u4e00\u6761\u53ef\u9a8c\u8bc1\u7684\u6839\u56e0\u94fe，\u9644 CLS \u62c9\u53d6\u811a\u672c\u3001\u7279\u5f81\u63d0\u53d6\u4ee3\u7801\u548c\u56de\u5f52\u9a8c\u8bc1\u65b9\u6cd5\u3002<\/p>\n

\n

0. \u5148\u8bf4\u7ed3\u8bba<\/h3>\n

\u5076\u53d1\u6545\u969c\u7684\u6392\u969c\u96be\u70b9\u4e0d\u5728"\u65e5\u5fd7\u4e0d\u591f"，\u800c\u5728"\u65e5\u5fd7\u592a\u591a\u4e14\u6ca1\u6709\u5bf9\u9f50\u7ef4\u5ea6"\u3002<\/p>\n

\u6b63\u786e\u7684\u987a\u5e8f\u662f：<\/p>\n

\u2460 \u522b\u5148\u770b\u65e5\u5fd7\u5185\u5bb9，\u5148\u7b97\u5206\u4f4d\u6570 \u2192 \u786e\u8ba4\u662f"\u5c3e\u90e8\u95ee\u9898"\u8fd8\u662f"\u6574\u4f53\u9000\u5316"
\n\u2461 \u522b\u62bd\u6837，\u8981"\u5168\u91cf\u6311\u51fa\u574f\u6837\u672c" \u2192 0.3% \u7684\u6982\u7387，\u62bd\u6837\u51e0\u4e4e\u5fc5\u7136\u62bd\u4e0d\u4e2d
\n\u2462 \u522b\u51ed\u5370\u8c61\u5f52\u56e0，\u8981\u505a"\u7279\u5f81\u5bf9\u6bd4\u8868" \u2192 \u597d\u6837\u672c vs \u574f\u6837\u672c，\u9010\u5b57\u6bb5\u6c42\u5dee\u5f02
\n\u2463 \u6700\u540e\u624d\u8ba9 AI \u4e0a\u573a \u2192 \u5b83\u7684\u4ef7\u503c\u662f\u628a 12 \u6761\u7279\u5f81\u5dee\u5f02\u62fc\u6210\u56e0\u679c\u94fe<\/p>\n

\u6309\u8fd9\u4e2a\u987a\u5e8f，\u6211\u4eec\u628a\u4e00\u4e2a"\u5076\u53d1 3s \u8d85\u65f6"\u7684\u95ee\u9898\u4ece 10.4 \u4e07\u884c\u65e5\u5fd7\u6536\u655b\u5230\u4e86 3 \u4e2a\u53ef\u9a8c\u8bc1\u7684\u5047\u8bbe，\u6700\u7ec8\u5b9a\u4f4d\u5230 \u6570\u636e\u5e93\u8fde\u63a5\u672a\u590d\u7528 + \u4e00\u6761\u67e5\u8be2\u672a\u8d70\u7d22\u5f15 + \u51b7\u542f\u52a8\u53e0\u52a0 \u4e09\u4e2a\u56e0\u7d20\u7684\u7ec4\u5408\u3002<\/p>\n

\u5bf9\u7167\u6570\u636e（\u4fee\u590d\u524d\u540e，\u5404\u89c2\u5bdf 7 \u5929）：<\/p>\n\n\u6307\u6807\u4fee\u590d\u524d\u4fee\u590d\u540e<\/tr>\n\n\n\n\n\n\n
\u8c03\u7528\u6b21\u6570<\/td>\n1,043,782<\/td>\n987,411<\/td>\n<\/tr>\n
\u8d85\u65f6\u6b21\u6570（>3s）<\/td>\n3,127<\/td>\n12<\/td>\n<\/tr>\n
\u8d85\u65f6\u7387<\/td>\n0.30%<\/td>\n0.0012%<\/td>\n<\/tr>\n
P99 \u8017\u65f6<\/td>\n3,180ms<\/td>\n412ms<\/td>\n<\/tr>\n
\u5e73\u5747\u8017\u65f6<\/td>\n178ms<\/td>\n164ms<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

\u6ce8\u610f\u6700\u540e\u4e00\u884c：\u5e73\u5747\u8017\u65f6\u51e0\u4e4e\u6ca1\u53d8\u3002 \u8fd9\u5c31\u662f\u4e3a\u4ec0\u4e48\u770b\u5747\u503c\u6c38\u8fdc\u627e\u4e0d\u5230\u95ee\u9898\u2014\u2014P99 \u6076\u5316\u4e86 7 \u500d，\u5747\u503c\u53ea\u52a8\u4e86 8%\u3002<\/p>\n

\n

1. \u95ee\u9898\u63cf\u8ff0\u4e0e\u4e3a\u4ec0\u4e48\u5e38\u89c4\u624b\u6bb5\u5931\u6548<\/h3>\n

\u73b0\u8c61：\u4e91\u51fd\u6570\u5076\u53d1\u8d85\u65f6\u3002\u5177\u4f53\u8868\u73b0\u662f：<\/p>\n

    \n
  • \u6bcf\u5929\u5927\u7ea6 300~500 \u6b21\u8c03\u7528\u8017\u65f6\u8d85\u8fc7 3 \u79d2（\u89e6\u53d1\u544a\u8b66\u9608\u503c）<\/li>\n
  • \u5360\u603b\u91cf\u7ea6 0.3%<\/li>\n
  • \u6ca1\u6709\u65f6\u95f4\u89c4\u5f8b：\u4e0d\u662f\u6574\u70b9\u3001\u4e0d\u662f\u6d41\u91cf\u9ad8\u5cf0\u3001\u548c\u53d1\u5e03\u65e0\u660e\u663e\u5173\u8054<\/li>\n
  • \u65e0\u6cd5\u590d\u73b0：\u624b\u52a8\u89e6\u53d1 500 \u6b21，\u4e00\u6b21\u90fd\u6ca1\u8d85\u65f6<\/li>\n<\/ul>\n

    \u4e09\u4e2a"\u6ca1\u6709"，\u76f4\u63a5\u6392\u9664\u4e86\u6700\u5e38\u89c1\u7684\u6392\u67e5\u8def\u5f84：<\/p>\n\n\u5e38\u89c4\u624b\u6bb5\u4e3a\u4ec0\u4e48\u5931\u6548<\/tr>\n\n\n\n\n\n
    \u770b\u76d1\u63a7\u5927\u76d8\u7684\u5747\u503c \/ \u6298\u7ebf<\/td>\n0.3% \u7684\u5f02\u5e38\u5bf9\u5747\u503c\u7684\u5f71\u54cd < 10%，\u88ab\u5b8c\u5168\u6df9\u6ca1<\/td>\n<\/tr>\n
    \u770b\u6d41\u91cf\u9ad8\u5cf0\u65f6\u6bb5<\/td>\n\u65e0\u89c4\u5f8b，\u4e0d\u662f\u6d41\u91cf\u95ee\u9898<\/td>\n<\/tr>\n
    \u968f\u673a\u62bd\u6837\u65e5\u5fd7<\/td>\n\u62bd 200 \u6761，\u9884\u671f\u547d\u4e2d 0.6 \u6761\u3002\u62bd\u4e0d\u4e2d<\/td>\n<\/tr>\n
    \u672c\u5730\/\u538b\u6d4b\u590d\u73b0<\/td>\n\u590d\u73b0\u6982\u7387 0.3%，\u9700\u8981\u8dd1 2000+ \u6b21\u624d\u6709\u8f83\u9ad8\u628a\u63e1，\u4e14\u73af\u5883\u5dee\u5f02\u5927<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

    \u7ed3\u8bba：\u5fc5\u987b\u4ece"\u5168\u91cf\u65e5\u5fd7\u4e2d\u7cbe\u786e\u6311\u51fa\u574f\u6837\u672c，\u7136\u540e\u505a\u5bf9\u6bd4\u5206\u6790"\u3002 \u8fd9\u5c31\u662f"\u8003\u53e4\u5f0f"\u2014\u2014\u4e0d\u53bb\u731c\u5893\u5728\u54ea\u91cc，\u800c\u662f\u628a\u6574\u7247\u5730\u626b\u4e00\u904d，\u5708\u51fa\u6240\u6709\u5f02\u5e38\u70b9\u4f4d，\u518d\u770b\u5b83\u4eec\u5171\u540c\u6709\u4ec0\u4e48\u3002<\/p>\n

    \n

    2. \u6570\u636e\u51c6\u5907：\u4ece CLS \u62c9\u65e5\u5fd7\u5230\u672c\u5730<\/h3>\n

    \u65e5\u5fd7\u901a\u8fc7\u4e91\u51fd\u6570\u65e5\u5fd7\u6295\u9012\u5230\u65e5\u5fd7\u670d\u52a1 CLS，\u7528 SearchLog \u63a5\u53e3\u6309\u65f6\u95f4\u7a97\u68c0\u7d22\u3002<\/p>\n

    # pull_cls.py<\/span>
    \nimport<\/span> json,<\/span> os,<\/span> time
    \nfrom<\/span> tencentcloud.<\/span>common import<\/span> credential
    \nfrom<\/span> tencentcloud.<\/span>common.<\/span>profile.<\/span>client_profile import<\/span> ClientProfile
    \nfrom<\/span> tencentcloud.<\/span>common.<\/span>profile.<\/span>http_profile import<\/span> HttpProfile
    \nfrom<\/span> tencentcloud.<\/span>cls.<\/span>v20201016 import<\/span> cls_client,<\/span> models<\/p>\n

    def<\/span> pull<\/span>(<\/span>topic_id:<\/span> str<\/span>,<\/span> start_ms:<\/span> int<\/span>,<\/span> end_ms:<\/span> int<\/span>,<\/span> out_path:<\/span> str<\/span>,<\/span> region=<\/span>"ap-guangzhou"<\/span>)<\/span>:<\/span>
    \n """
    \n start_ms \/ end_ms: \u6beb\u79d2\u65f6\u95f4\u6233
    \n Query \u8bed\u6cd5\u53c2\u8003 CLS \u68c0\u7d22\u8bed\u6cd5，\u4f8b\u5982:
    \n "REPORT" \u6216 '"timeout"' \u6216 \u6309\u5b57\u6bb5\u8fc7\u6ee4 'funcName:"orderQuery"'
    \n """<\/span>
    \n cred =<\/span> credential.<\/span>Credential(<\/span>
    \n os.<\/span>environ[<\/span>"TENCENTCLOUD_SECRET_ID"<\/span>]<\/span>,<\/span>
    \n os.<\/span>environ[<\/span>"TENCENTCLOUD_SECRET_KEY"<\/span>]<\/span>,<\/span>
    \n )<\/span>
    \n hp =<\/span> HttpProfile(<\/span>endpoint=<\/span>"cls.tencentcloudapi.com"<\/span>,<\/span> reqTimeout=<\/span>60<\/span>)<\/span>
    \n client =<\/span> cls_client.<\/span>ClsClient(<\/span>cred,<\/span> region,<\/span> ClientProfile(<\/span>httpProfile=<\/span>hp)<\/span>)<\/span><\/p>\n

    written =<\/span> 0<\/span>
    \n with<\/span> open<\/span>(<\/span>out_path,<\/span> "w"<\/span>,<\/span> encoding=<\/span>"utf-8"<\/span>)<\/span> as<\/span> f:<\/span>
    \n # \u6309\u5c0f\u65f6\u5207\u7247，\u907f\u514d\u5355\u6b21\u8fd4\u56de\u4e0a\u9650\u622a\u65ad<\/span>
    \n step =<\/span> 3600<\/span> *<\/span> 1000<\/span>
    \n t =<\/span> start_ms
    \n while<\/span> t <<\/span> end_ms:<\/span>
    \n req =<\/span> models.<\/span>SearchLogRequest(<\/span>)<\/span>
    \n req.<\/span>TopicId =<\/span> topic_id
    \n req.<\/span>From =<\/span> t
    \n req.<\/span>To =<\/span> min<\/span>(<\/span>t +<\/span> step,<\/span> end_ms)<\/span>
    \n req.<\/span>Query =<\/span> "*"<\/span> # \u5148\u5168\u91cf\u62c9，\u7b5b\u9009\u653e\u672c\u5730\u505a，\u907f\u514d\u4e8c\u6b21\u62c9\u53d6<\/span>
    \n req.<\/span>Limit =<\/span> 1000<\/span>
    \n req.<\/span>Sort =<\/span> "asc"<\/span>
    \n req.<\/span>SyntaxRule =<\/span> 1<\/span> # \u6309\u4f60\u8d26\u53f7\u7684\u68c0\u7d22\u8bed\u6cd5\u7248\u672c\u8bbe\u7f6e<\/span><\/p>\n

    resp =<\/span> client.<\/span>SearchLog(<\/span>req)<\/span>
    \n data =<\/span> json.<\/span>loads(<\/span>resp.<\/span>to_json_string(<\/span>)<\/span>)<\/span>
    \n for<\/span> r in<\/span> data.<\/span>get(<\/span>"Results"<\/span>,<\/span> [<\/span>]<\/span>)<\/span>:<\/span>
    \n f.<\/span>write(<\/span>json.<\/span>dumps(<\/span>{<\/span>
    \n "time"<\/span>:<\/span> r.<\/span>get(<\/span>"Time"<\/span>)<\/span>,<\/span>
    \n "source"<\/span>:<\/span> r.<\/span>get(<\/span>"Source"<\/span>)<\/span>,<\/span>
    \n "raw"<\/span>:<\/span> r.<\/span>get(<\/span>"RawLog"<\/span>)<\/span>,<\/span>
    \n }<\/span>,<\/span> ensure_ascii=<\/span>False<\/span>)<\/span> +<\/span> "\\\\n"<\/span>)<\/span>
    \n written +=<\/span> 1<\/span>
    \n if<\/span> not<\/span> data.<\/span>get(<\/span>"ListOver"<\/span>)<\/span>:<\/span>
    \n # \u8be5\u65f6\u95f4\u7247\u8fd8\u6709\u66f4\u591a\u6570\u636e，\u7f29\u5c0f\u7a97\u53e3\u91cd\u8bd5<\/span>
    \n step =<\/span> max<\/span>(<\/span>60<\/span> *<\/span> 1000<\/span>,<\/span> step \/\/<\/span> 2<\/span>)<\/span>
    \n continue<\/span>
    \n t +=<\/span> 3600<\/span> *<\/span> 1000<\/span>
    \n step =<\/span> 3600<\/span> *<\/span> 1000<\/span>
    \n time.<\/span>sleep(<\/span>0.1<\/span>)<\/span>
    \n print<\/span>(<\/span>f"pulled <\/span>{<\/span>written}<\/span><\/span> log lines -> <\/span>{<\/span>out_path}<\/span><\/span>"<\/span><\/span>)<\/span><\/p>\n

    \u91c7\u96c6\u5c42\u4e09\u4e2a\u8981\u70b9：<\/p>\n

  • \u4e0d\u8981\u5e26\u4e1a\u52a1\u8fc7\u6ee4\u6761\u4ef6\u53bb\u62c9\u3002 \u5148\u7528 * \u5168\u91cf\u62c9，\u7b5b\u9009\u5728\u672c\u5730\u505a\u3002\u7406\u7531\u5f88\u7b80\u5355：\u4f60\u5982\u679c\u5df2\u7ecf\u77e5\u9053\u8be5\u600e\u4e48\u7b5b，\u4f60\u5c31\u5df2\u7ecf\u77e5\u9053\u7b54\u6848\u4e86\u3002\u800c\u4e14"\u597d\u6837\u672c"\u4e5f\u9700\u8981\u62c9\u4e0b\u6765\u505a\u5bf9\u6bd4\u2014\u2014\u53ea\u62c9\u574f\u6837\u672c，\u4f60\u6ca1\u6cd5\u505a\u7279\u5f81\u5bf9\u6bd4\u3002<\/li>\n
  • \u6309\u5c0f\u65f6\u5207\u7247 + ListOver \u5224\u65ad\u622a\u65ad\u3002 \u5355\u6b21\u8fd4\u56de\u6709\u4e0a\u9650，\u8d85\u5927\u65f6\u95f4\u7a97\u4f1a\u9759\u9ed8\u622a\u65ad\u3002\u5224\u65ad ListOver \u5e76\u5728\u622a\u65ad\u65f6\u7f29\u5c0f\u7a97\u53e3\u91cd\u8bd5，\u53ef\u4ee5\u907f\u514d"\u6f0f\u6389\u6700\u5173\u952e\u7684\u90a3\u6761\u65e5\u5fd7"\u8fd9\u79cd\u6700\u8ba9\u4eba\u5d29\u6e83\u7684\u60c5\u51b5\u3002<\/li>\n
  • \u843d\u76d8\u4fdd\u7559\u539f\u59cb\u884c\u3002 \u7ed3\u6784\u5316\u89e3\u6790\u653e\u5728\u5206\u6790\u9636\u6bb5\u505a\u3002\u539f\u59cb\u65e5\u5fd7\u53ea\u62c9\u4e00\u6b21，\u4e4b\u540e\u6240\u6709\u5b9e\u9a8c\u90fd\u5728\u672c\u5730\u91cd\u8dd1，\u907f\u514d\u53cd\u590d\u8c03\u7528\u63a5\u53e3\u3002<\/li>\n
    \n

    3. \u6309 RequestId \u805a\u5408\u7269\u5316：\u8fd9\u4e00\u6b65\u51b3\u5b9a\u6210\u8d25<\/h3>\n

    \u65e5\u5fd7\u662f"\u884c"\u7684\u96c6\u5408，\u4f46\u4e00\u6b21\u8c03\u7528\u624d\u662f\u6392\u969c\u5355\u4f4d\u3002\u5fc5\u987b\u5148\u805a\u5408\u7269\u5316，\u624d\u80fd\u505a\u5bf9\u6bd4\u3002<\/p>\n

    # materialize.py<\/span>
    \nimport<\/span> json,<\/span> re
    \nfrom<\/span> collections import<\/span> defaultdict<\/p>\n

    # \u5404\u5b57\u6bb5\u7684\u6b63\u5219\/\u53d6\u503c\u903b\u8f91\u9700\u8981\u6309\u4f60\u5b9e\u9645\u7684\u65e5\u5fd7\u683c\u5f0f\u8c03\u6574<\/span>
    \nRE_REQUEST_ID =<\/span> re.<\/span>compile<\/span>(<\/span>r'"request_id"\\\\s*:\\\\s*"([^"]+)"'<\/span>)<\/span>
    \nRE_DURATION =<\/span> re.<\/span>compile<\/span>(<\/span>r'"duration_ms"\\\\s*:\\\\s*([\\\\d.]+)'<\/span>)<\/span><\/p>\n

    def<\/span> extract_fields<\/span>(<\/span>raw:<\/span> str<\/span>)<\/span> –<\/span>><\/span> dict<\/span>:<\/span>
    \n """
    \n \u4e91\u51fd\u6570\u5e73\u53f0\u65e5\u5fd7\u7684\u5173\u952e\u5b57\u6bb5\u901a\u5e38\u51fa\u73b0\u5728\u7ed3\u6784\u5316\u7684\u5e73\u53f0\u65e5\u5fd7\u4e2d；
    \n \u4e1a\u52a1\u65e5\u5fd7\u5219\u53d6\u51b3\u4e8e\u4f60\u81ea\u5df1\u7684\u6253\u70b9\u65b9\u5f0f\u3002\u4ee5\u4e0b\u4e3a\u901a\u7528\u515c\u5e95\u7b56\u7565\u3002
    \n """<\/span>
    \n fields =<\/span> {<\/span>"requestId"<\/span>:<\/span> None<\/span>,<\/span> "durationMs"<\/span>:<\/span> None<\/span>,<\/span> "coldStart"<\/span>:<\/span> None<\/span>,<\/span>
    \n "dbQueryMs"<\/span>:<\/span> None<\/span>,<\/span> "dbConnMs"<\/span>:<\/span> None<\/span>,<\/span> "retryCount"<\/span>:<\/span> None<\/span>,<\/span>
    \n "memoryMb"<\/span>:<\/span> None<\/span>,<\/span> "concurrency"<\/span>:<\/span> None<\/span>,<\/span> "userTag"<\/span>:<\/span> None<\/span>}<\/span><\/p>\n

    m =<\/span> RE_REQUEST_ID.<\/span>search(<\/span>raw)<\/span>
    \n if<\/span> m:<\/span>
    \n fields[<\/span>"requestId"<\/span>]<\/span> =<\/span> m.<\/span>group(<\/span>1<\/span>)<\/span>
    \n m =<\/span> RE_DURATION.<\/span>search(<\/span>raw)<\/span>
    \n if<\/span> m:<\/span>
    \n fields[<\/span>"durationMs"<\/span>]<\/span> =<\/span> float<\/span>(<\/span>m.<\/span>group(<\/span>1<\/span>)<\/span>)<\/span><\/p>\n

    # \u4e1a\u52a1\u6253\u70b9：\u5efa\u8bae\u7edf\u4e00\u6210 JSON \u65e5\u5fd7，\u4e00\u884c\u4e00\u6761<\/span>
    \n try<\/span>:<\/span>
    \n obj =<\/span> json.<\/span>loads(<\/span>raw)<\/span>
    \n for<\/span> src,<\/span> dst in<\/span> (<\/span>(<\/span>"duration_ms"<\/span>,<\/span> "durationMs"<\/span>)<\/span>,<\/span> (<\/span>"cold_start"<\/span>,<\/span> "coldStart"<\/span>)<\/span>,<\/span>
    \n (<\/span>"db_query_ms"<\/span>,<\/span> "dbQueryMs"<\/span>)<\/span>,<\/span> (<\/span>"db_conn_ms"<\/span>,<\/span> "dbConnMs"<\/span>)<\/span>,<\/span>
    \n (<\/span>"retry_count"<\/span>,<\/span> "retryCount"<\/span>)<\/span>,<\/span> (<\/span>"memory_mb"<\/span>,<\/span> "memoryMb"<\/span>)<\/span>,<\/span>
    \n (<\/span>"user_tag"<\/span>,<\/span> "userTag"<\/span>)<\/span>)<\/span>:<\/span>
    \n if<\/span> src in<\/span> obj:<\/span>
    \n fields[<\/span>dst]<\/span> =<\/span> obj[<\/span>src]<\/span>
    \n except<\/span> Exception:<\/span>
    \n pass<\/span>
    \n return<\/span> fields<\/p>\n

    def<\/span> materialize<\/span>(<\/span>in_path:<\/span> str<\/span>)<\/span> –<\/span>><\/span> dict<\/span>:<\/span>
    \n """\u6309 requestId \u628a\u6563\u843d\u7684\u65e5\u5fd7\u884c\u805a\u6210"\u4e00\u6b21\u8c03\u7528"\u7684\u753b\u50cf"""<\/span>
    \n calls =<\/span> defaultdict(<\/span>lambda<\/span>:<\/span> {<\/span>"durMs"<\/span>:<\/span> None<\/span>,<\/span> "lines"<\/span>:<\/span> 0<\/span>,<\/span> "cold"<\/span>:<\/span> False<\/span>,<\/span>
    \n "dbQueryMs"<\/span>:<\/span> 0.0<\/span>,<\/span> "dbConnMs"<\/span>:<\/span> 0.0<\/span>,<\/span>
    \n "retry"<\/span>:<\/span> 0<\/span>,<\/span> "tags"<\/span>:<\/span> {<\/span>}<\/span>}<\/span>)<\/span>
    \n with<\/span> open<\/span>(<\/span>in_path,<\/span> encoding=<\/span>"utf-8"<\/span>)<\/span> as<\/span> f:<\/span>
    \n for<\/span> line in<\/span> f:<\/span>
    \n rec =<\/span> json.<\/span>loads(<\/span>line)<\/span>
    \n raw =<\/span> rec.<\/span>get(<\/span>"raw"<\/span>,<\/span> ""<\/span>)<\/span>
    \n fs =<\/span> extract_fields(<\/span>raw)<\/span>
    \n rid =<\/span> fs[<\/span>"requestId"<\/span>]<\/span>
    \n if<\/span> not<\/span> rid:<\/span>
    \n continue<\/span>
    \n c =<\/span> calls[<\/span>rid]<\/span>
    \n c[<\/span>"lines"<\/span>]<\/span> +=<\/span> 1<\/span>
    \n if<\/span> fs[<\/span>"durationMs"<\/span>]<\/span>:<\/span>
    \n c[<\/span>"durMs"<\/span>]<\/span> =<\/span> max<\/span>(<\/span>c[<\/span>"durMs"<\/span>]<\/span> or<\/span> 0<\/span>,<\/span> fs[<\/span>"durationMs"<\/span>]<\/span>)<\/span>
    \n if<\/span> fs[<\/span>"coldStart"<\/span>]<\/span>:<\/span>
    \n c[<\/span>"cold"<\/span>]<\/span> =<\/span> True<\/span>
    \n for<\/span> k,<\/span> dst in<\/span> (<\/span>(<\/span>"dbQueryMs"<\/span>,<\/span> "dbQueryMs"<\/span>)<\/span>,<\/span> (<\/span>"dbConnMs"<\/span>,<\/span> "dbConnMs"<\/span>)<\/span>)<\/span>:<\/span>
    \n if<\/span> fs[<\/span>k]<\/span>:<\/span>
    \n c[<\/span>dst]<\/span> +=<\/span> fs[<\/span>k]<\/span>
    \n if<\/span> fs[<\/span>"retryCount"<\/span>]<\/span>:<\/span>
    \n c[<\/span>"retry"<\/span>]<\/span> =<\/span> max<\/span>(<\/span>c[<\/span>"retry"<\/span>]<\/span>,<\/span> int<\/span>(<\/span>fs[<\/span>"retryCount"<\/span>]<\/span>)<\/span>)<\/span>
    \n for<\/span> t in<\/span> (<\/span>"userTag"<\/span>,<\/span> "memoryMb"<\/span>,<\/span> "concurrency"<\/span>)<\/span>:<\/span>
    \n if<\/span> fs[<\/span>t]<\/span> is<\/span> not<\/span> None<\/span>:<\/span>
    \n c[<\/span>"tags"<\/span>]<\/span>[<\/span>t]<\/span> =<\/span> fs[<\/span>t]<\/span>
    \n return<\/span> dict<\/span>(<\/span>calls)<\/span><\/p>\n

    \u5982\u679c\u65e5\u5fd7\u91cc\u6ca1\u6709 requestId \u600e\u4e48\u529e？ \u4e09\u79cd\u515c\u5e95\u7b56\u7565，\u6309\u53ef\u9760\u6027\u6392\u5e8f：<\/p>\n

  • \u6309\u8c03\u7528\u65e5\u5fd7\u4e2d\u7684\u552f\u4e00\u4e1a\u52a1\u6807\u8bc6\u805a\u5408（\u5982\u8ba2\u5355\u53f7 + \u65f6\u95f4\u7a97）\u3002\u6700\u53ef\u9760，\u524d\u63d0\u662f\u4e1a\u52a1\u6709\u7a33\u5b9a\u6807\u8bc6\u3002<\/li>\n
  • \u6309\u65f6\u95f4\u7a97\u53e3 + \u51fd\u6570\u540d\u805a\u5408\u3002\u628a\u540c\u4e00\u79d2\u5185\u540c\u51fd\u6570\u7684\u65e5\u5fd7\u884c\u5f52\u4e3a\u4e00\u6b21\u8c03\u7528\u3002\u5e76\u53d1\u4f4e\u65f6\u53ef\u7528，\u5e76\u53d1\u9ad8\u65f6\u4f1a\u8bef\u5408\u5e76\u3002<\/li>\n
  • \u6539\u9020\u6253\u70b9，\u8865\u4e0a requestId\u3002\u5982\u679c\u4ee5\u4e0a\u90fd\u4e0d\u53ef\u884c，\u90a3\u8fd9\u6b21\u7684\u7ed3\u8bba\u53ea\u80fd\u662f"\u6253\u70b9\u4e0d\u8db3"，\u5148\u628a\u6253\u70b9\u8865\u4e0a\u2014\u2014\u8fd9\u4e0d\u662f\u59a5\u534f，\u8fd9\u662f\u6b63\u786e\u7684\u4e0b\u4e00\u6b65\u3002<\/li>\n

    \u8fd9\u4e00\u6b65\u6700\u5927\u7684\u4ef7\u503c\u5728\u4e8e：\u628a"10 \u4e07\u884c\u65e5\u5fd7"\u53d8\u6210"1 \u4e07\u6b21\u8c03\u7528"\u3002 \u7ef4\u5ea6\u4e00\u6362，\u95ee\u9898\u5c31\u4ece"\u6587\u672c\u5206\u6790"\u53d8\u6210\u4e86"\u6837\u672c\u5bf9\u6bd4"，\u540e\u8005\u624d\u662f\u53ef\u8ba1\u7b97\u7684\u3002<\/p>\n

    \n

    4. \u7b97\u5206\u4f4d\u6570，\u786e\u8ba4\u662f"\u5c3e\u90e8\u95ee\u9898"<\/h3>\n

    # quantify.py<\/span>
    \nimport<\/span> numpy as<\/span> np<\/p>\n

    def<\/span> quantiles<\/span>(<\/span>calls:<\/span> dict<\/span>)<\/span> –<\/span>><\/span> dict<\/span>:<\/span>
    \n dur =<\/span> np.<\/span>array(<\/span>[<\/span>c[<\/span>"durMs"<\/span>]<\/span> for<\/span> c in<\/span> calls.<\/span>values(<\/span>)<\/span> if<\/span> c[<\/span>"durMs"<\/span>]<\/span>]<\/span>,<\/span> dtype=<\/span>float<\/span>)<\/span>
    \n return<\/span> {<\/span>
    \n "n"<\/span>:<\/span> len<\/span>(<\/span>dur)<\/span>,<\/span>
    \n "mean"<\/span>:<\/span> round<\/span>(<\/span>float<\/span>(<\/span>dur.<\/span>mean(<\/span>)<\/span>)<\/span>,<\/span> 1<\/span>)<\/span>,<\/span>
    \n "p50"<\/span>:<\/span> round<\/span>(<\/span>float<\/span>(<\/span>np.<\/span>percentile(<\/span>dur,<\/span> 50<\/span>)<\/span>)<\/span>,<\/span> 1<\/span>)<\/span>,<\/span>
    \n "p90"<\/span>:<\/span> round<\/span>(<\/span>float<\/span>(<\/span>np.<\/span>percentile(<\/span>dur,<\/span> 90<\/span>)<\/span>)<\/span>,<\/span> 1<\/span>)<\/span>,<\/span>
    \n "p99"<\/span>:<\/span> round<\/span>(<\/span>float<\/span>(<\/span>np.<\/span>percentile(<\/span>dur,<\/span> 99<\/span>)<\/span>)<\/span>,<\/span> 1<\/span>)<\/span>,<\/span>
    \n "p999"<\/span>:<\/span> round<\/span>(<\/span>float<\/span>(<\/span>np.<\/span>percentile(<\/span>dur,<\/span> 99.9<\/span>)<\/span>)<\/span>,<\/span> 1<\/span>)<\/span>,<\/span>
    \n "max"<\/span>:<\/span> round<\/span>(<\/span>float<\/span>(<\/span>dur.<\/span>max<\/span>(<\/span>)<\/span>)<\/span>,<\/span> 1<\/span>)<\/span>,<\/span>
    \n "over_3s"<\/span>:<\/span> int<\/span>(<\/span>(<\/span>dur ><\/span> 3000<\/span>)<\/span>.<\/span>sum<\/span>(<\/span>)<\/span>)<\/span>,<\/span>
    \n "over_3s_ratio"<\/span>:<\/span> round<\/span>(<\/span>float<\/span>(<\/span>(<\/span>dur ><\/span> 3000<\/span>)<\/span>.<\/span>mean(<\/span>)<\/span>)<\/span>,<\/span> 5<\/span>)<\/span>,<\/span>
    \n }<\/span><\/p>\n

    \u5b9e\u6d4b\u8f93\u51fa：<\/p>\n

    {<\/span>"n"<\/span>:<\/span> 1043782<\/span>,<\/span> "mean"<\/span>:<\/span> 178.3<\/span>,<\/span> "p50"<\/span>:<\/span> 142.0<\/span>,<\/span> "p90"<\/span>:<\/span> 265.0<\/span>,<\/span>
    \n "p99"<\/span>:<\/span> 3180.0<\/span>,<\/span> "p999"<\/span>:<\/span> 4102.0<\/span>,<\/span> "max"<\/span>:<\/span> 12844.0<\/span>,<\/span>
    \n "over_3s"<\/span>:<\/span> 3127<\/span>,<\/span> "over_3s_ratio"<\/span>:<\/span> 0.003<\/span>}<\/span><\/p>\n

    \u600e\u4e48\u8bfb\u8fd9\u7ec4\u6570：<\/p>\n

      \n
    • p50 = 142ms \u8bf4\u660e\u7edd\u5927\u591a\u6570\u8bf7\u6c42\u5f88\u5feb \u2192 \u4e0d\u662f\u6574\u4f53\u6027\u80fd\u9000\u5316，\u662f\u5c11\u6570\u6837\u672c\u52a3\u5316\u3002<\/li>\n
    • p99 \/ p50 = 22.4\u3002\u8fd9\u4e2a\u6bd4\u503c\u6b63\u5e38\u5e94\u5728 5~10 \u4e4b\u95f4（\u5e38\u89c1 Web \u670d\u52a1\u7684\u7ecf\u9a8c\u533a\u95f4）\u300222 \u500d\u8bf4\u660e\u5c3e\u90e8\u4e25\u91cd\u62d6\u5c3e\u3002<\/li>\n
    • p999 = 4102 \u4e0e p99 = 3180 \u63a5\u8fd1，max = 12844 \u662f\u79bb\u7fa4\u70b9\u3002<\/li>\n
    • mean \u53ea\u6bd4 p50 \u9ad8 25%，\u8fdb\u4e00\u6b65\u786e\u8ba4：\u5747\u503c\u88ab\u5c3e\u90e8\u62c9\u52a8\u7684\u5e45\u5ea6\u5f88\u5c0f，\u6240\u4ee5\u5927\u76d8\u770b\u4e0d\u51fa\u6765\u3002<\/li>\n<\/ul>\n

      \u5230\u8fd9\u4e00\u6b65，"\u8fd9\u662f\u4e2a\u5c3e\u90e8\u95ee\u9898"\u5c31\u88ab\u6570\u636e\u786e\u8ba4\u4e86\u3002\u63a5\u4e0b\u6765\u4e0d\u53bb\u770b\u5747\u503c，\u4e0d\u53bb\u770b\u5927\u76d8，\u53ea\u505a\u4e00\u4ef6\u4e8b：\u628a 3127 \u4e2a\u574f\u6837\u672c\u548c 104 \u4e07\u4e2a\u597d\u6837\u672c\u505a\u5bf9\u6bd4\u3002<\/p>\n

      \n

      5. \u7279\u5f81\u5bf9\u6bd4\u8868：\u8ba9\u6570\u636e\u66ff\u4f60\u5f52\u56e0<\/h3>\n

      # profile.py<\/span>
      \nimport<\/span> numpy as<\/span> np
      \nfrom<\/span> collections import<\/span> Counter<\/p>\n

      BAD_THRESHOLD_MS =<\/span> 3000<\/span><\/p>\n

      def<\/span> profile<\/span>(<\/span>calls:<\/span> dict<\/span>)<\/span> –<\/span>><\/span> dict<\/span>:<\/span>
      \n good,<\/span> bad =<\/span> [<\/span>]<\/span>,<\/span> [<\/span>]<\/span>
      \n for<\/span> c in<\/span> calls.<\/span>values(<\/span>)<\/span>:<\/span>
      \n if<\/span> not<\/span> c[<\/span>"durMs"<\/span>]<\/span>:<\/span>
      \n continue<\/span>
      \n (<\/span>bad if<\/span> c[<\/span>"durMs"<\/span>]<\/span> ><\/span> BAD_THRESHOLD_MS else<\/span> good)<\/span>.<\/span>append(<\/span>c)<\/span><\/p>\n

      def<\/span> agg<\/span>(<\/span>group,<\/span> key,<\/span> is_bool=<\/span>False<\/span>)<\/span>:<\/span>
      \n if<\/span> is_bool:<\/span>
      \n return<\/span> round<\/span>(<\/span>sum<\/span>(<\/span>1<\/span> for<\/span> g in<\/span> group if<\/span> g[<\/span>key]<\/span>)<\/span> \/<\/span> max<\/span>(<\/span>len<\/span>(<\/span>group)<\/span>,<\/span> 1<\/span>)<\/span>,<\/span> 4<\/span>)<\/span>
      \n vals =<\/span> [<\/span>g[<\/span>key]<\/span> for<\/span> g in<\/span> group if<\/span> g.<\/span>get(<\/span>key)<\/span>]<\/span>
      \n return<\/span> round<\/span>(<\/span>float<\/span>(<\/span>np.<\/span>mean(<\/span>vals)<\/span>)<\/span>,<\/span> 2<\/span>)<\/span> if<\/span> vals else<\/span> None<\/span><\/p>\n

      def<\/span> pct<\/span>(<\/span>group,<\/span> key,<\/span> target)<\/span>:<\/span>
      \n if<\/span> not<\/span> group:<\/span>
      \n return<\/span> None<\/span>
      \n cnt =<\/span> sum<\/span>(<\/span>1<\/span> for<\/span> g in<\/span> group if<\/span> g[<\/span>"tags"<\/span>]<\/span>.<\/span>get(<\/span>key)<\/span> ==<\/span> target)<\/span>
      \n return<\/span> round<\/span>(<\/span>cnt \/<\/span> len<\/span>(<\/span>group)<\/span>,<\/span> 4<\/span>)<\/span><\/p>\n

      return<\/span> {<\/span>
      \n "\u6837\u672c\u6570"<\/span>:<\/span> {<\/span>"good"<\/span>:<\/span> len<\/span>(<\/span>good)<\/span>,<\/span> "bad"<\/span>:<\/span> len<\/span>(<\/span>bad)<\/span>}<\/span>,<\/span>
      \n "\u51b7\u542f\u52a8\u5360\u6bd4"<\/span>:<\/span> {<\/span>"good"<\/span>:<\/span> agg(<\/span>good,<\/span> "cold"<\/span>,<\/span> True<\/span>)<\/span>,<\/span> "bad"<\/span>:<\/span> agg(<\/span>bad,<\/span> "cold"<\/span>,<\/span> True<\/span>)<\/span>}<\/span>,<\/span>
      \n "\u5e73\u5747DB\u67e5\u8be2\u8017\u65f6(ms)"<\/span>:<\/span> {<\/span>"good"<\/span>:<\/span> agg(<\/span>good,<\/span> "dbQueryMs"<\/span>)<\/span>,<\/span> "bad"<\/span>:<\/span> agg(<\/span>bad,<\/span> "dbQueryMs"<\/span>)<\/span>}<\/span>,<\/span>
      \n "\u5e73\u5747DB\u8fde\u63a5\u8017\u65f6(ms)"<\/span>:<\/span> {<\/span>"good"<\/span>:<\/span> agg(<\/span>good,<\/span> "dbConnMs"<\/span>)<\/span>,<\/span> "bad"<\/span>:<\/span> agg(<\/span>bad,<\/span> "dbConnMs"<\/span>)<\/span>}<\/span>,<\/span>
      \n "\u5e73\u5747\u91cd\u8bd5\u6b21\u6570"<\/span>:<\/span> {<\/span>"good"<\/span>:<\/span> agg(<\/span>good,<\/span> "retry"<\/span>)<\/span>,<\/span> "bad"<\/span>:<\/span> agg(<\/span>bad,<\/span> "retry"<\/span>)<\/span>}<\/span>,<\/span>
      \n "\u603b\u65e5\u5fd7\u884c\u6570\/\u8c03\u7528"<\/span>:<\/span> {<\/span>"good"<\/span>:<\/span> round<\/span>(<\/span>len<\/span>(<\/span>sum<\/span>(<\/span>[<\/span>g[<\/span>"lines"<\/span>]<\/span> for<\/span> g in<\/span> good]<\/span>,<\/span> [<\/span>]<\/span>)<\/span>)<\/span> \/<\/span> max<\/span>(<\/span>len<\/span>(<\/span>good)<\/span>,<\/span> 1<\/span>)<\/span>,<\/span> 2<\/span>)<\/span>,<\/span>
      \n "bad"<\/span>:<\/span> round<\/span>(<\/span>len<\/span>(<\/span>sum<\/span>(<\/span>[<\/span>g[<\/span>"lines"<\/span>]<\/span> for<\/span> g in<\/span> bad]<\/span>,<\/span> [<\/span>]<\/span>)<\/span>)<\/span> \/<\/span> max<\/span>(<\/span>len<\/span>(<\/span>bad)<\/span>,<\/span> 1<\/span>)<\/span>,<\/span> 2<\/span>)<\/span>}<\/span>,<\/span>
      \n }<\/span><\/p>\n

      \u5b9e\u6d4b\u5bf9\u6bd4\u8868（\u8fd9\u662f\u6574\u4e2a\u6392\u969c\u8fc7\u7a0b\u91cc\u4fe1\u606f\u91cf\u6700\u5927\u7684\u4e00\u5f20\u8868）：<\/p>\n\n\u7279\u5f81\u597d\u6837\u672c (n=1,040,655)\u574f\u6837\u672c (n=3,127)\u500d\u6570<\/tr>\n\n\n\n\n\n\n\n
      \u51b7\u542f\u52a8\u5360\u6bd4<\/td>\n0.8%<\/td>\n11.4%<\/td>\n14x<\/td>\n<\/tr>\n
      \u5e73\u5747 DB \u67e5\u8be2\u8017\u65f6<\/td>\n21 ms<\/td>\n312 ms<\/td>\n15x<\/td>\n<\/tr>\n
      \u5e73\u5747 DB \u8fde\u63a5\u8017\u65f6<\/td>\n3 ms<\/td>\n268 ms<\/td>\n89x<\/td>\n<\/tr>\n
      \u5e73\u5747\u91cd\u8bd5\u6b21\u6570<\/td>\n0.02<\/td>\n1.41<\/td>\n70x<\/td>\n<\/tr>\n
      \u5185\u5b58\u5cf0\u503c<\/td>\n128 MB<\/td>\n219 MB<\/td>\n1.7x<\/td>\n<\/tr>\n
      \u7528\u6237\u5206\u5c42<\/td>\n\u65e0\u660e\u663e\u5dee\u5f02<\/td>\n\u65e0\u660e\u663e\u5dee\u5f02<\/td>\n\u2014<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

      \u4e09\u6761\u7279\u5f81\u5dee\u5f02\u6781\u5176\u663e\u8457（14x \/ 89x \/ 70x），\u4e24\u6761\u65e0\u5dee\u5f02\u3002 \u8fd9\u5c31\u662f\u8003\u53e4\u5f0f\u6392\u969c\u7684\u6838\u5fc3\u4ea7\u51fa\u2014\u2014\u5b83\u628a"\u5076\u53d1\u8d85\u65f6"\u8fd9\u4e2a\u7b3c\u7edf\u7684\u95ee\u9898，\u538b\u7f29\u6210\u4e86\u4e09\u4e2a\u5177\u4f53\u4e14\u53ef\u9a8c\u8bc1\u7684\u7591\u70b9：<\/p>\n

    • DB \u8fde\u63a5\u8017\u65f6 89 \u500d \u2192 \u8fde\u63a5\u5efa\u7acb\u8fc7\u7a0b\u6709\u95ee\u9898，\u9ad8\u5ea6\u6000\u7591\u6ca1\u6709\u590d\u7528\u8fde\u63a5\u6c60<\/li>\n
    • DB \u67e5\u8be2\u8017\u65f6 15 \u500d \u2192 \u67e5\u8be2\u672c\u8eab\u4e5f\u6709\u95ee\u9898，\u6000\u7591\u672a\u8d70\u7d22\u5f15 \/ \u6709\u5168\u8868\u626b\u63cf<\/li>\n
    • \u51b7\u542f\u52a8\u5360\u6bd4 14 \u500d \u2192 \u51b7\u542f\u52a8\u662f"\u653e\u5927\u5668"\u800c\u4e0d\u662f"\u6839\u56e0"（\u51b7\u542f\u52a8\u672c\u8eab\u4e0d\u4ea7\u751f 3 \u79d2\u8017\u65f6）<\/li>\n

      \u800c\u4e14\u5b83\u6392\u9664\u4e86\u4e24\u6761\u5f88\u5e38\u89c1\u7684\u731c\u6d4b：\u7528\u6237\u5206\u5c42\u65e0\u5dee\u5f02（\u4e0d\u662f\u7279\u5b9a\u7528\u6237\u89e6\u53d1），\u5185\u5b58\u65e0\u663e\u8457\u5dee\u5f02（\u4e0d\u662f OOM \u8fb9\u7f18）\u3002<\/p>\n

      \n

      6. \u4ea4\u7ed9 AI \u7684\u90a3\u4e00\u6b65：\u62fc\u56e0\u679c\u94fe<\/h3>\n

      \u5230\u8fd9\u91cc，\u4eba\u5de5\u7684\u6210\u672c\u5df2\u7ecf\u5f88\u4f4e\u4e86\u3002\u5269\u4e0b\u7684\u5de5\u4f5c\u662f\u628a\u4e09\u6761\u5b64\u7acb\u7279\u5f81\u62fc\u6210\u4e00\u6761\u56e0\u679c\u94fe，\u5e76\u7ed9\u51fa\u9a8c\u8bc1\u65b9\u6848\u2014\u2014\u8fd9\u6b63\u662f AI \u64c5\u957f\u7684\u90e8\u5206\u3002<\/p>\n

      \u5582\u7ed9\u52a9\u624b\u7684\u5185\u5bb9\u4e0d\u9700\u8981 10 \u4e07\u884c\u65e5\u5fd7，\u53ea\u9700\u8981：\u5206\u4f4d\u6570\u8868 + \u7279\u5f81\u5bf9\u6bd4\u8868 + \u53ef\u7591\u6837\u672c\u7684\u5b8c\u6574\u8c03\u7528\u94fe（\u6311 5 \u6761）\u3002<\/p>\n

      \u4f60\u662f\u6027\u80fd\u6392\u969c\u52a9\u624b\u3002\u4e0b\u9762\u662f\u4e00\u6b21\u5076\u53d1\u8d85\u65f6\u6545\u969c\u7684\u4e09\u4efd\u8bc1\u636e\u3002<\/p>\n

      \u3010\u7ea6\u675f\u3011
      \n1. \u53ea\u5141\u8bb8\u57fa\u4e8e\u7ed9\u5b9a\u6570\u636e\u63a8\u7406，\u4e0d\u5f97\u5f15\u5165"\u5e38\u89c1\u505a\u6cd5"\u5f0f\u7684\u901a\u7528\u731c\u6d4b\u3002
      \n2. \u6bcf\u4e2a\u7ed3\u8bba\u5fc5\u987b\u6807\u6ce8\u5b83\u4f9d\u8d56\u7684\u662f\u54ea\u51e0\u4e2a\u6570\u636e\u70b9\u3002
      \n3. \u5fc5\u987b\u7ed9\u51fa"\u8bc1\u4f2a\u65b9\u6848"：\u5982\u679c\u6211\u54ea\u4e2a\u5047\u8bbe\u9519\u4e86，\u600e\u4e48\u7528\u6570\u636e\u9a8c\u8bc1\u51fa\u6765\u3002
      \n4. \u8f93\u51fa JSON\u3002<\/p>\n

      \u3010\u8bc1\u636e 1：\u5206\u4f4d\u6570\u3011
      \np50=142ms p90=265ms p99=3180ms p999=4102ms max=12844ms
      \nmean=178ms, \u8d85\u65f6\u7387 0.30%（3127\/1043782）<\/p>\n

      \u3010\u8bc1\u636e 2：\u597d\u574f\u6837\u672c\u7279\u5f81\u5bf9\u6bd4\u3011
      \n\u51b7\u542f\u52a8\u5360\u6bd4 0.8% vs 11.4%
      \nDB\u67e5\u8be2\u8017\u65f6 21ms vs 312ms
      \nDB\u8fde\u63a5\u8017\u65f6 3ms vs 268ms
      \n\u91cd\u8bd5\u6b21\u6570 0.02 vs 1.41<\/p>\n

      \u3010\u8bc1\u636e 3：3 \u6761\u574f\u6837\u672c\u7684\u5b8c\u6574\u8c03\u7528\u94fe\u3011
      \n（\u6b64\u5904\u7c98\u8d34\u8c03\u7528\u94fe，\u542b\u5404\u9636\u6bb5\u8017\u65f6\u6253\u70b9）<\/p>\n

      \u3010\u8981\u6c42\u8f93\u51fa\u3011
      \n{
      \n "rootCauseChain": [{"order":1,"cause":"…","evidence":["…"],"contribution":"…"}],
      \n "amplifiers": [{"factor":"…","evidence":["…"]}],
      \n "ruledOut": [{"hypothesis":"…","why":"…"}],
      \n "verifications": [{"hypothesis":"…","method":"…","expected":"…","falsifyIf":"…"}],
      \n "fixPriority": [{"priority":1,"fix":"…","expectedGain":"…"}]
      \n}<\/p>\n

      AI \u7ed9\u51fa\u7684\u8f93\u51fa（\u8282\u9009）：<\/p>\n

      {<\/span>
      \n "rootCauseChain"<\/span>:<\/span> [<\/span>
      \n {<\/span>
      \n "order"<\/span>:<\/span> 1<\/span>,<\/span>
      \n "cause"<\/span>:<\/span> "\u6570\u636e\u5e93\u8fde\u63a5\u672a\u590d\u7528：\u6bcf\u6b21\u8c03\u7528\u65b0\u5efa\u8fde\u63a5，\u5728\u7279\u5b9a\u7f51\u7edc\/\u8d1f\u8f7d\u6761\u4ef6\u4e0b\u63e1\u624b\u88ab\u62c9\u957f\u5230 250ms+"<\/span>,<\/span>
      \n "evidence"<\/span>:<\/span> [<\/span>"DB\u8fde\u63a5\u8017\u65f6 3ms vs 268ms，89\u500d\u5dee\u5f02"<\/span>,<\/span> "\u91cd\u8bd5\u6b21\u6570 0.02 vs 1.41：\u8fde\u63a5\u5931\u8d25\u540e\u91cd\u8bd5，\u8fdb\u4e00\u6b65\u62c9\u957f\u603b\u8017\u65f6"<\/span>]<\/span>,<\/span>
      \n "contribution"<\/span>:<\/span> "\u5355\u6b21\u8c03\u7528\u76f4\u63a5\u589e\u52a0 265ms \u57fa\u7ebf，\u4e14\u91cd\u8bd5\u4f7f\u589e\u76ca\u7ffb\u500d"<\/span>
      \n }<\/span>,<\/span>
      \n {<\/span>
      \n "order"<\/span>:<\/span> 2<\/span>,<\/span>
      \n "cause"<\/span>:<\/span> "\u90e8\u5206\u67e5\u8be2\u672a\u8d70\u7d22\u5f15，\u5728\u8fde\u63a5\u5efa\u7acb\u8f83\u6162\u65f6\u53e0\u52a0\u5168\u8868\u626b\u63cf"<\/span>,<\/span>
      \n "evidence"<\/span>:<\/span> [<\/span>"DB\u67e5\u8be2\u8017\u65f6 21ms vs 312ms，15\u500d\u5dee\u5f02"<\/span>,<\/span> "\u5dee\u5206\u8bf4\u660e\u67e5\u8be2\u5c42\u72ec\u7acb\u5b58\u5728\u52a3\u5316，\u4e0d\u662f\u7eaf\u8fde\u63a5\u95ee\u9898"<\/span>]<\/span>,<\/span>
      \n "contribution"<\/span>:<\/span> "\u5728\u8fde\u63a5\u6162\u7684\u57fa\u7840\u4e0a\u518d\u53e0\u52a0\u7ea6 290ms"<\/span>
      \n }<\/span>
      \n ]<\/span>,<\/span>
      \n "amplifiers"<\/span>:<\/span> [<\/span>
      \n {<\/span>
      \n "factor"<\/span>:<\/span> "\u51b7\u542f\u52a8"<\/span>,<\/span>
      \n "evidence"<\/span>:<\/span> [<\/span>"\u51b7\u542f\u52a8\u5360\u6bd4 0.8% vs 11.4%：\u51b7\u542f\u52a8\u4f7f\u8c03\u7528\u843d\u5165"<\/span>\u8fde\u63a5\u91cd\u5efa"\u8def\u5f84\u7684\u6982\u7387\u4e0a\u5347，\u672c\u8eab\u4e0d\u76f4\u63a5\u9020\u6210 3s \u8017\u65f6"<\/span>]<\/span>
      \n }<\/span>
      \n ]<\/span>,<\/span>
      \n "ruledOut"<\/span>:<\/span> [<\/span>
      \n {<\/span> "hypothesis"<\/span>:<\/span> "\u7279\u5b9a\u7528\u6237\u89e6\u53d1\u7684\u5927\u6570\u636e\u91cf\u8bf7\u6c42"<\/span>,<\/span> "why"<\/span>:<\/span> "\u7528\u6237\u5206\u5c42\u65e0\u5dee\u5f02"<\/span> }<\/span>,<\/span>
      \n {<\/span> "hypothesis"<\/span>:<\/span> "\u5185\u5b58\u4e0d\u8db3\u5bfc\u81f4 GC \u6296\u52a8\u6216 OOM \u8fb9\u7f18"<\/span>,<\/span> "why"<\/span>:<\/span> "\u5185\u5b58\u5cf0\u503c\u4ec5 1.7 \u500d，\u65e0\u663e\u8457\u5dee\u5f02"<\/span> }<\/span>
      \n ]<\/span>,<\/span>
      \n "verifications"<\/span>:<\/span> [<\/span>
      \n {<\/span>
      \n "hypothesis"<\/span>:<\/span> "\u8fde\u63a5\u672a\u590d\u7528"<\/span>,<\/span>
      \n "method"<\/span>:<\/span> "\u7edf\u8ba1\u574f\u6837\u672c\u4e2d db_conn_ms > 100ms \u7684\u8c03\u7528，\u5176\u4e0a\u6e38\u662f\u5426\u90fd\u4f34\u968f"<\/span>\u65b0\u5efa\u8fde\u63a5"\u65e5\u5fd7"<\/span>,<\/span>
      \n "expected"<\/span>:<\/span> "\u9ad8\u4e8e 90% \u7684\u574f\u6837\u672c\u5e94\u547d\u4e2d\u8be5\u6a21\u5f0f"<\/span>,<\/span>
      \n "falsifyIf"<\/span>:<\/span> "\u82e5\u547d\u4e2d\u7387\u4f4e\u4e8e 50%，\u5219\u8fde\u63a5\u8017\u65f6\u662f\u7ed3\u679c\u800c\u975e\u539f\u56e0"<\/span>
      \n }<\/span>,<\/span>
      \n {<\/span>
      \n "hypothesis"<\/span>:<\/span> "\u67e5\u8be2\u672a\u8d70\u7d22\u5f15"<\/span>,<\/span>
      \n "method"<\/span>:<\/span> "\u5bf9\u574f\u6837\u672c\u4e2d\u7684 SQL \u505a EXPLAIN，\u68c0\u67e5 type=ALL \u7684\u6bd4\u4f8b"<\/span>,<\/span>
      \n "expected"<\/span>:<\/span> "\u574f\u6837\u672c SQL \u7684 type=ALL \u6bd4\u4f8b\u663e\u8457\u9ad8\u4e8e\u597d\u6837\u672c"<\/span>,<\/span>
      \n "falsifyIf"<\/span>:<\/span> "\u82e5\u4e24\u8005\u7d22\u5f15\u4f7f\u7528\u60c5\u51b5\u4e00\u81f4，\u5219\u67e5\u8be2\u5c42\u52a3\u5316\u5c5e\u8fde\u5e26\u6548\u5e94"<\/span>
      \n }<\/span>
      \n ]<\/span>,<\/span>
      \n "fixPriority"<\/span>:<\/span> [<\/span>
      \n {<\/span> "priority"<\/span>:<\/span> 1<\/span>,<\/span> "fix"<\/span>:<\/span> "\u6539\u4e3a\u51fd\u6570\u5916\u521d\u59cb\u5316\u5ba2\u6237\u7aef\/\u542f\u7528\u8fde\u63a5\u6c60\u5e76\u590d\u7528"<\/span>,<\/span> "expectedGain"<\/span>:<\/span> "\u6d88\u9664 265ms \u57fa\u7ebf\u589e\u91cf，\u8d85\u65f6\u7387\u9884\u8ba1\u964d\u81f3 5% \u4ee5\u5185"<\/span> }<\/span>,<\/span>
      \n {<\/span> "priority"<\/span>:<\/span> 2<\/span>,<\/span> "fix"<\/span>:<\/span> "\u8865\u7d22\u5f15 + \u4e3a\u6162\u67e5\u8be2\u52a0\u8d85\u65f6"<\/span>,<\/span> "expectedGain"<\/span>:<\/span> "\u6d88\u9664 290ms \u53e0\u52a0"<\/span> }<\/span>
      \n ]<\/span>
      \n}<\/span><\/p>\n

      \u6ce8\u610f\u8fd9\u6bb5\u8f93\u51fa\u91cc\u6700\u6709\u4ef7\u503c\u7684\u90e8\u5206\u4e0d\u662f rootCauseChain，\u800c\u662f verifications \u548c ruledOut\u3002 \u524d\u8005\u544a\u8bc9\u4f60"\u5982\u679c\u4f60\u9519\u4e86，\u600e\u4e48\u7528\u6570\u636e\u53d1\u73b0"，\u540e\u8005\u544a\u8bc9\u4f60"\u54ea\u4e9b\u770b\u8d77\u6765\u5408\u7406\u7684\u731c\u6d4b\u5df2\u7ecf\u88ab\u6570\u636e\u6392\u9664，\u522b\u518d\u82b1\u65f6\u95f4\u53bb\u67e5"\u3002\u8fd9\u4e24\u9879\u624d\u662f\u8ba9 AI \u5f52\u56e0\u53ef\u4fe1\u7684\u5173\u952e\u2014\u2014\u5b83\u5fc5\u987b\u628a\u81ea\u5df1\u653e\u5230\u53ef\u88ab\u8bc1\u4f2a\u7684\u4f4d\u7f6e\u4e0a\u3002<\/p>\n

      \n

      7. \u9a8c\u8bc1\u4e0e\u4fee\u590d<\/h3>\n

      \u6309 AI \u7ed9\u51fa\u7684\u8bc1\u4f2a\u65b9\u6848，\u6211\u4eec\u8dd1\u4e86\u4e24\u6761\u9a8c\u8bc1：<\/p>\n

      \u9a8c\u8bc1 1：\u574f\u6837\u672c\u662f\u5426\u90fd\u4f34\u968f"\u65b0\u5efa\u8fde\u63a5"<\/p>\n

      # verify_conn.py<\/span>
      \nbad_with_new_conn =<\/span> 0<\/span>
      \nbad_total =<\/span> 0<\/span>
      \nfor<\/span> c in<\/span> bad_calls:<\/span>
      \n bad_total +=<\/span> 1<\/span>
      \n if<\/span> any<\/span>(<\/span>"connection established"<\/span> in<\/span> l for<\/span> l in<\/span> c[<\/span>"raw_lines"<\/span>]<\/span>[<\/span>:<\/span>3<\/span>]<\/span>)<\/span>:<\/span>
      \n bad_with_new_conn +=<\/span> 1<\/span>
      \nprint<\/span>(<\/span>f"<\/span>{<\/span>bad_with_new_conn}<\/span><\/span>\/<\/span>{<\/span>bad_total}<\/span><\/span> = <\/span>{<\/span>bad_with_new_conn\/<\/span>bad_total:<\/span>.2%<\/span>}<\/span><\/span>"<\/span><\/span>)<\/span>
      \n# \u8f93\u51fa: 2918\/3127 = 93.31%<\/span><\/p>\n

      93.31% > \u9884\u671f\u7684 90%，\u5047\u8bbe 1 \u6210\u7acb\u3002<\/p>\n

      \u9a8c\u8bc1 2：\u574f\u6837\u672c SQL \u7684\u7d22\u5f15\u4f7f\u7528\u60c5\u51b5<\/p>\n

      \u5bf9\u574f\u6837\u672c\u4e2d\u63d0\u53d6\u51fa\u7684 47 \u6761\u552f\u4e00 SQL \u505a EXPLAIN：<\/p>\n\n\u7c7b\u578b\u6761\u6570\u5360\u6bd4<\/tr>\n\n\n\n\n
      type=ALL（\u5168\u8868\u626b\u63cf）<\/td>\n11<\/td>\n23.4%<\/td>\n<\/tr>\n
      type=range<\/td>\n28<\/td>\n59.6%<\/td>\n<\/tr>\n
      type=ref\/eq_ref<\/td>\n8<\/td>\n17.0%<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

      \u597d\u6837\u672c SQL \u7684 type=ALL \u5360\u6bd4\u4e3a 1.2%\u300223.4% vs 1.2%，\u5047\u8bbe 2 \u6210\u7acb\u3002<\/p>\n

      \u4fee\u590d\u52a8\u4f5c：<\/p>\n

    • \u628a\u6570\u636e\u5e93\u5ba2\u6237\u7aef\u521d\u59cb\u5316\u79fb\u5230\u51fd\u6570handler \u5916（\u590d\u7528），\u5e76\u663e\u5f0f\u914d\u7f6e\u8fde\u63a5\u6c60<\/li>\n
    • \u4e3a 11 \u6761\u5168\u8868\u626b\u63cf SQL \u8865\u7d22\u5f15<\/li>\n
    • \u4e3a DB \u67e5\u8be2\u52a0\u8d85\u65f6（\u8d85\u8fc7 800ms \u76f4\u63a5\u5931\u8d25\u5e76\u964d\u7ea7），\u907f\u514d\u62d6\u5230 3s<\/li>\n
    • \u8fde\u63a5\u5931\u8d25\u4e0d\u518d\u76f2\u76ee\u91cd\u8bd5，\u6539\u4e3a\u5e26\u9000\u907f\u7684 1 \u6b21\u91cd\u8bd5<\/li>\n

      \u56de\u5f52\u9a8c\u8bc1（\u89c2\u5bdf 7 \u5929）：<\/p>\n\n\u6307\u6807\u4fee\u590d\u524d\u4fee\u590d\u540e<\/tr>\n\n\n\n\n\n\n
      \u8d85\u65f6\u6b21\u6570<\/td>\n3,127<\/td>\n12<\/td>\n<\/tr>\n
      \u8d85\u65f6\u7387<\/td>\n0.30%<\/td>\n0.0012%<\/td>\n<\/tr>\n
      P99<\/td>\n3,180ms<\/td>\n412ms<\/td>\n<\/tr>\n
      P999<\/td>\n4,102ms<\/td>\n638ms<\/td>\n<\/tr>\n
      \u5e73\u5747\u8017\u65f6<\/td>\n178ms<\/td>\n164ms<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n
      \n

      8. \u65b9\u6cd5\u8bba\u63d0\u70bc：\u8003\u53e4\u5f0f\u6392\u969c\u7684 5 \u6761\u539f\u5219<\/h3>\n
    • \u5148\u91cf\u5206\u4f4d\u6570，\u518d\u770b\u5185\u5bb9\u3002 p99\/p50 \u7684\u6bd4\u503c\u80fd\u5728 30 \u79d2\u5185\u544a\u8bc9\u4f60"\u662f\u5c3e\u90e8\u95ee\u9898\u8fd8\u662f\u6574\u4f53\u9000\u5316"，\u800c\u8fd9\u4e2a\u5224\u65ad\u51b3\u5b9a\u4e86\u4f60\u540e\u9762\u6240\u6709\u5de5\u4f5c\u7684\u65b9\u5411\u3002\u6bd4\u503c >15 \u5c31\u4e0d\u8981\u518d\u76ef\u5747\u503c\u4e86\u3002<\/li>\n
    • \u574f\u6837\u672c\u8981\u5168\u91cf\u6311，\u4e0d\u8981\u62bd\u6837\u3002 \u62bd\u6837\u9002\u5408"\u9a8c\u8bc1\u5df2\u77e5\u89c4\u5f8b"，\u4e0d\u9002\u5408"\u53d1\u73b0\u672a\u77e5\u89c4\u5f8b"\u30020.3% \u7684\u95ee\u9898，\u62bd\u6837\u547d\u4e2d\u7684\u6982\u7387\u548c\u4e70\u5f69\u7968\u5dee\u4e0d\u591a\u3002<\/li>\n
    • \u5bf9\u6bd4\u8868\u6bd4\u65e5\u5fd7\u5185\u5bb9\u66f4\u6709\u4fe1\u606f\u91cf\u3002 \u4e00\u5f20"\u597d vs \u574f"\u7684\u7279\u5f81\u5bf9\u6bd4\u8868，\u4ef7\u503c\u9ad8\u4e8e\u4efb\u4f55\u4e00\u6bb5\u65e5\u5fd7\u539f\u6587\u3002\u5dee\u5f02\u500d\u6570\u5927\u7684\u5b57\u6bb5\u5c31\u662f\u5acc\u7591\u72af，\u65e0\u5dee\u5f02\u7684\u5b57\u6bb5\u5c31\u662f\u4e0d\u5728\u573a\u8bc1\u660e\u3002<\/li>\n
    • \u8ba9 AI \u8f93\u51fa"\u8bc1\u4f2a\u65b9\u6848"，\u800c\u4e0d\u662f"\u7ed3\u8bba"\u3002 \u4e00\u4e2a\u4e0d\u53ef\u8bc1\u4f2a\u7684\u7ed3\u8bba\u5728\u6392\u969c\u91cc\u7b49\u4e8e\u96f6\u3002\u8981\u6c42\u5b83\u7ed9\u51fa"\u5982\u679c\u6211\u9519\u4e86，\u6570\u636e\u4f1a\u600e\u4e48\u663e\u793a"，\u8fd9\u624d\u628a AI \u4ece"\u731c"\u53d8\u6210\u4e86"\u63d0\u5047\u8bbe"\u3002<\/li>\n
    • \u628a"\u653e\u5927\u5668"\u548c"\u6839\u56e0"\u5206\u5f00\u3002 \u51b7\u542f\u52a8\u3001\u91cd\u8bd5\u3001GC \u8fd9\u4e9b\u901a\u5e38\u662f\u653e\u5927\u5668\u2014\u2014\u5b83\u4eec\u8ba9\u95ee\u9898\u4ece 100ms \u53d8\u6210 3s，\u4f46\u5b83\u4eec\u4e0d\u662f\u8d77\u70b9\u3002\u6df7\u6dc6\u8fd9\u4e24\u8005\u4f1a\u5bfc\u81f4\u4f60\u82b1\u5927\u529b\u6c14\u4f18\u5316\u51b7\u542f\u52a8，\u800c\u8d85\u65f6\u4f9d\u65e7\u3002<\/li>\n

      \u6700\u540e\u4e00\u6761\u5c24\u5176\u91cd\u8981：\u5728\u8fd9\u6b21\u7684\u6848\u4f8b\u91cc，\u51b7\u542f\u52a8\u7684\u5dee\u5f02\u500d\u6570\u662f\u6700\u5927\u7684\u90a3\u7c7b（14x）\u4e4b\u4e00，\u4f46\u5b83\u4e0d\u662f\u6839\u56e0\u3002 \u5982\u679c\u53ea\u6309\u5dee\u5f02\u500d\u6570\u6392\u5e8f\u53bb\u4fee，\u4f60\u4f1a\u5148\u53bb\u4f18\u5316\u51b7\u542f\u52a8，\u7136\u540e\u53d1\u73b0\u8d85\u65f6\u7387\u51e0\u4e4e\u6ca1\u964d\u3002\u5dee\u5f02\u500d\u6570\u8bf4\u660e\u7684\u662f"\u76f8\u5173\u6027"，\u56e0\u679c\u94fe\u53ea\u80fd\u9760\u673a\u5236\u63a8\u7406 + \u8bc1\u4f2a\u9a8c\u8bc1\u5f97\u5230\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"

      \u5b9e\u6218｜10 \u4e07\u884c\u4e91\u51fd\u6570\u65e5\u5fd7\u7684\\”\u8003\u53e4\u5f0f\\”\u6392\u969c：\u7528\u817e\u8baf\u4e91\u52a9\u624b\u628a\u5076\u53d1\u8d85\u65f6\u538b\u7f29\u6210\u4e00\u6761\u6839\u56e0\u94fe\u5206\u7c7b\u4e13\u680f：\u817e\u8baf\u4e91\u52a9\u624b\u5b9e\u6218 \/ \u53ef\u89c2\u6d4b\u6027
      \n\u6807\u7b7e：\u817e\u8baf\u4e91\u52a9\u624b\u3001\u4e91\u51fd\u6570 SCF\u3001\u65e5\u5fd7\u670d\u52a1 CLS\u3001\u5076\u53d1\u6545\u969c\u3001P99\u3001\u6839\u56e0\u5206\u6790\u3001Python
      \n\u6458\u8981：\u5076\u53d1\u8d85\u65f6（0.3% \u6982\u7387）\u662f\u6392\u969c\u91cc\u6700\u96be\u7684\u4e00\u7c7b：\u56fe\u8868\u770b\u5747\u503c\u770b\u4e0d\u51fa\u6765，\u62bd\u6837\u53c8\u62bd\u4e0d\u4e2d\u3002\u672c\u6587\u7ed9\u51fa\u300c\u65e5\u5fd7\u805a\u5408 \u2192 \u5c3e\u90e8\u6837\u672c\u753b\u50cf \u2192 \u7279\u5f81\u5bf9\u6bd4\u8868<\/p>\n","protected":false},"author":2,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[50,351],"topic":[],"class_list":["post-108235","post","type-post","status-publish","format-standard","hentry","category-server","tag-50","tag-351"],"yoast_head":"\n\u5b9e\u6218\uff5c10 \u4e07\u884c\u4e91\u51fd\u6570\u65e5\u5fd7\u7684\u201c\u8003\u53e4\u5f0f\u201c\u6392\u969c\uff1a\u7528\u817e\u8baf\u4e91\u52a9\u624b\u628a\u5076\u53d1\u8d85\u65f6\u538b\u7f29\u6210\u4e00\u6761\u6839\u56e0\u94fe - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.wsisp.com\/helps\/108235.html\" \/>\n<meta property=\"og:locale\" content=\"zh_CN\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"\u5b9e\u6218\uff5c10 \u4e07\u884c\u4e91\u51fd\u6570\u65e5\u5fd7\u7684\u201c\u8003\u53e4\u5f0f\u201c\u6392\u969c\uff1a\u7528\u817e\u8baf\u4e91\u52a9\u624b\u628a\u5076\u53d1\u8d85\u65f6\u538b\u7f29\u6210\u4e00\u6761\u6839\u56e0\u94fe - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\" \/>\n<meta property=\"og:description\" content=\"\u5b9e\u6218｜10 \u4e07\u884c\u4e91\u51fd\u6570\u65e5\u5fd7\u7684"\u8003\u53e4\u5f0f"\u6392\u969c：\u7528\u817e\u8baf\u4e91\u52a9\u624b\u628a\u5076\u53d1\u8d85\u65f6\u538b\u7f29\u6210\u4e00\u6761\u6839\u56e0\u94fe\u5206\u7c7b\u4e13\u680f：\u817e\u8baf\u4e91\u52a9\u624b\u5b9e\u6218 \/ \u53ef\u89c2\u6d4b\u6027 \u6807\u7b7e：\u817e\u8baf\u4e91\u52a9\u624b\u3001\u4e91\u51fd\u6570 SCF\u3001\u65e5\u5fd7\u670d\u52a1 CLS\u3001\u5076\u53d1\u6545\u969c\u3001P99\u3001\u6839\u56e0\u5206\u6790\u3001Python \u6458\u8981：\u5076\u53d1\u8d85\u65f6（0.3% \u6982\u7387）\u662f\u6392\u969c\u91cc\u6700\u96be\u7684\u4e00\u7c7b：\u56fe\u8868\u770b\u5747\u503c\u770b\u4e0d\u51fa\u6765，\u62bd\u6837\u53c8\u62bd\u4e0d\u4e2d\u3002\u672c\u6587\u7ed9\u51fa\u300c\u65e5\u5fd7\u805a\u5408 \u2192 \u5c3e\u90e8\u6837\u672c\u753b\u50cf \u2192 \u7279\u5f81\u5bf9\u6bd4\u8868\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.wsisp.com\/helps\/108235.html\" \/>\n<meta property=\"og:site_name\" content=\"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\" \/>\n<meta property=\"article:published_time\" content=\"2026-09-21T03:24:42+00:00\" \/>\n<meta name=\"author\" content=\"admin\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"\u4f5c\u8005\" \/>\n\t<meta name=\"twitter:data1\" content=\"admin\" \/>\n\t<meta name=\"twitter:label2\" content=\"\u9884\u8ba1\u9605\u8bfb\u65f6\u95f4\" \/>\n\t<meta name=\"twitter:data2\" content=\"7 \u5206\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/108235.html\",\"url\":\"https:\/\/www.wsisp.com\/helps\/108235.html\",\"name\":\"\u5b9e\u6218\uff5c10 \u4e07\u884c\u4e91\u51fd\u6570\u65e5\u5fd7\u7684\u201c\u8003\u53e4\u5f0f\u201c\u6392\u969c\uff1a\u7528\u817e\u8baf\u4e91\u52a9\u624b\u628a\u5076\u53d1\u8d85\u65f6\u538b\u7f29\u6210\u4e00\u6761\u6839\u56e0\u94fe - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\",\"isPartOf\":{\"@id\":\"https:\/\/www.wsisp.com\/helps\/#website\"},\"datePublished\":\"2026-09-21T03:24:42+00:00\",\"dateModified\":\"2026-09-21T03:24:42+00:00\",\"author\":{\"@id\":\"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41\"},\"breadcrumb\":{\"@id\":\"https:\/\/www.wsisp.com\/helps\/108235.html#breadcrumb\"},\"inLanguage\":\"zh-Hans\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/www.wsisp.com\/helps\/108235.html\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/108235.html#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"\u9996\u9875\",\"item\":\"https:\/\/www.wsisp.com\/helps\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"\u5b9e\u6218\uff5c10 \u4e07\u884c\u4e91\u51fd\u6570\u65e5\u5fd7\u7684\u201c\u8003\u53e4\u5f0f\u201c\u6392\u969c\uff1a\u7528\u817e\u8baf\u4e91\u52a9\u624b\u628a\u5076\u53d1\u8d85\u65f6\u538b\u7f29\u6210\u4e00\u6761\u6839\u56e0\u94fe\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/#website\",\"url\":\"https:\/\/www.wsisp.com\/helps\/\",\"name\":\"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\",\"description\":\"\u9999\u6e2f\u670d\u52a1\u5668_\u9999\u6e2f\u4e91\u670d\u52a1\u5668\u8d44\u8baf_\u670d\u52a1\u5668\u5e2e\u52a9\u6587\u6863_\u670d\u52a1\u5668\u6559\u7a0b\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/www.wsisp.com\/helps\/?s={search_term_string}\"},\"query-input\":\"required name=search_term_string\"}],\"inLanguage\":\"zh-Hans\"},{\"@type\":\"Person\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41\",\"name\":\"admin\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"zh-Hans\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery\",\"contentUrl\":\"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery\",\"caption\":\"admin\"},\"sameAs\":[\"http:\/\/wp.wsisp.com\"],\"url\":\"https:\/\/www.wsisp.com\/helps\/author\/admin\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"\u5b9e\u6218\uff5c10 \u4e07\u884c\u4e91\u51fd\u6570\u65e5\u5fd7\u7684\u201c\u8003\u53e4\u5f0f\u201c\u6392\u969c\uff1a\u7528\u817e\u8baf\u4e91\u52a9\u624b\u628a\u5076\u53d1\u8d85\u65f6\u538b\u7f29\u6210\u4e00\u6761\u6839\u56e0\u94fe - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.wsisp.com\/helps\/108235.html","og_locale":"zh_CN","og_type":"article","og_title":"\u5b9e\u6218\uff5c10 \u4e07\u884c\u4e91\u51fd\u6570\u65e5\u5fd7\u7684\u201c\u8003\u53e4\u5f0f\u201c\u6392\u969c\uff1a\u7528\u817e\u8baf\u4e91\u52a9\u624b\u628a\u5076\u53d1\u8d85\u65f6\u538b\u7f29\u6210\u4e00\u6761\u6839\u56e0\u94fe - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","og_description":"\u5b9e\u6218｜10 \u4e07\u884c\u4e91\u51fd\u6570\u65e5\u5fd7\u7684\"\u8003\u53e4\u5f0f\"\u6392\u969c：\u7528\u817e\u8baf\u4e91\u52a9\u624b\u628a\u5076\u53d1\u8d85\u65f6\u538b\u7f29\u6210\u4e00\u6761\u6839\u56e0\u94fe\u5206\u7c7b\u4e13\u680f：\u817e\u8baf\u4e91\u52a9\u624b\u5b9e\u6218 \/ \u53ef\u89c2\u6d4b\u6027 \u6807\u7b7e：\u817e\u8baf\u4e91\u52a9\u624b\u3001\u4e91\u51fd\u6570 SCF\u3001\u65e5\u5fd7\u670d\u52a1 CLS\u3001\u5076\u53d1\u6545\u969c\u3001P99\u3001\u6839\u56e0\u5206\u6790\u3001Python \u6458\u8981：\u5076\u53d1\u8d85\u65f6（0.3% \u6982\u7387）\u662f\u6392\u969c\u91cc\u6700\u96be\u7684\u4e00\u7c7b：\u56fe\u8868\u770b\u5747\u503c\u770b\u4e0d\u51fa\u6765，\u62bd\u6837\u53c8\u62bd\u4e0d\u4e2d\u3002\u672c\u6587\u7ed9\u51fa\u300c\u65e5\u5fd7\u805a\u5408 \u2192 \u5c3e\u90e8\u6837\u672c\u753b\u50cf \u2192 \u7279\u5f81\u5bf9\u6bd4\u8868","og_url":"https:\/\/www.wsisp.com\/helps\/108235.html","og_site_name":"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","article_published_time":"2026-09-21T03:24:42+00:00","author":"admin","twitter_card":"summary_large_image","twitter_misc":{"\u4f5c\u8005":"admin","\u9884\u8ba1\u9605\u8bfb\u65f6\u95f4":"7 \u5206"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/www.wsisp.com\/helps\/108235.html","url":"https:\/\/www.wsisp.com\/helps\/108235.html","name":"\u5b9e\u6218\uff5c10 \u4e07\u884c\u4e91\u51fd\u6570\u65e5\u5fd7\u7684\u201c\u8003\u53e4\u5f0f\u201c\u6392\u969c\uff1a\u7528\u817e\u8baf\u4e91\u52a9\u624b\u628a\u5076\u53d1\u8d85\u65f6\u538b\u7f29\u6210\u4e00\u6761\u6839\u56e0\u94fe - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","isPartOf":{"@id":"https:\/\/www.wsisp.com\/helps\/#website"},"datePublished":"2026-09-21T03:24:42+00:00","dateModified":"2026-09-21T03:24:42+00:00","author":{"@id":"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41"},"breadcrumb":{"@id":"https:\/\/www.wsisp.com\/helps\/108235.html#breadcrumb"},"inLanguage":"zh-Hans","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.wsisp.com\/helps\/108235.html"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/www.wsisp.com\/helps\/108235.html#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"\u9996\u9875","item":"https:\/\/www.wsisp.com\/helps"},{"@type":"ListItem","position":2,"name":"\u5b9e\u6218\uff5c10 \u4e07\u884c\u4e91\u51fd\u6570\u65e5\u5fd7\u7684\u201c\u8003\u53e4\u5f0f\u201c\u6392\u969c\uff1a\u7528\u817e\u8baf\u4e91\u52a9\u624b\u628a\u5076\u53d1\u8d85\u65f6\u538b\u7f29\u6210\u4e00\u6761\u6839\u56e0\u94fe"}]},{"@type":"WebSite","@id":"https:\/\/www.wsisp.com\/helps\/#website","url":"https:\/\/www.wsisp.com\/helps\/","name":"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","description":"\u9999\u6e2f\u670d\u52a1\u5668_\u9999\u6e2f\u4e91\u670d\u52a1\u5668\u8d44\u8baf_\u670d\u52a1\u5668\u5e2e\u52a9\u6587\u6863_\u670d\u52a1\u5668\u6559\u7a0b","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.wsisp.com\/helps\/?s={search_term_string}"},"query-input":"required name=search_term_string"}],"inLanguage":"zh-Hans"},{"@type":"Person","@id":"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41","name":"admin","image":{"@type":"ImageObject","inLanguage":"zh-Hans","@id":"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/image\/","url":"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery","contentUrl":"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery","caption":"admin"},"sameAs":["http:\/\/wp.wsisp.com"],"url":"https:\/\/www.wsisp.com\/helps\/author\/admin"}]}},"_links":{"self":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/posts\/108235","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/comments?post=108235"}],"version-history":[{"count":0,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/posts\/108235\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/media?parent=108235"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/categories?post=108235"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/tags?post=108235"},{"taxonomy":"topic","embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/topic?post=108235"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}