我有一个在Tomcat上运行的GWT应用程序，最近我发布了一个新版本的应用程序，突然我的服务器响应时间从平均500ms到平均15s。 我已经运行了我所知道的所有监控工具。
mysqltuner输出
[OK] Logged in using credentials from debian maintenance account. -------- General Statistics -------------------------------------------------- [--] Skipped version check for MySQLTuner script [OK] Currently running supported MySQL version 5.1.63-0ubuntu0.10.04.1-log [OK] Operating on 64-bit architecture -------- Storage Engine Statistics ------------------------------------------- [--] Status: +Archive -BDB -Federated +InnoDB -ISAM -NDBCluster [--] Data in MyISAM tables: 370M (Tables: 52) [--] Data in InnoDB tables: 697M (Tables: 1749) [!!] Total fragmented tables: 1754 -------- Security Recommendations ------------------------------------------- [OK] All database users have passwords assigned -------- Performance Metrics ------------------------------------------------- [--] Up for: 19h 25m 41s (1M q [28.122 qps], 1K conn, TX: 2B, RX: 1B) [--] Reads / Writes: 98% / 2% [--] Total buffers: 1.0G global + 2.7M per thread (500 max threads) [OK] Maximum possible memory usage: 2.4G (30% of installed RAM) [OK] Slow queries: 0% (1/1M) [OK] Highest usage of available connections: 34% (173/500) [OK] Key buffer size / total MyISAM indexes: 16.0M/279.0K [OK] Key buffer hit rate: 99.9% (50K cached / 40 reads) [OK] Query cache efficiency: 61.4% (844K cached / 1M selects) [!!] Query cache prunes per day: 553779 [OK] Sorts requiring temporary tables: 0% (0 temp sorts / 34K sorts) [OK] Temporary tables created on disk: 4% (4K on disk / 102K total) [OK] Thread cache hit rate: 84% (185 created / 1K connections) [!!] Table cache hit rate: 0% (256 open / 27K opened) [OK] Open file limit used: 0% (20/2K) [OK] Table locks acquired immediately: 100% (692K immediate / 692K locks) [OK] InnoDB data size / buffer pool: 697.2M/1.0G -------- Recommendations ----------------------------------------------------- General recommendations: Run OPTIMIZE TABLE to defragment tables for better performance MySQL started within last 24 hours - recommendations may be inaccurate Enable the slow query log to troubleshoot bad queries Increase table_cache gradually to avoid file descriptor limits Variables to adjust: query_cache_size (> 16M) table_cache (> 256) 如果有人感兴趣，问题是我们在某个时候超过了连接池中的最大连接数，这只会导致线程中的雪球效应等待连接池。 这就解释了为什么机器“睡着了”，所有的东西都在队列中排队等待连接。