targetRuntime: Used to specify the runtime target for �→ generated code. It has MyBatis3DynamicSql / MyBatis3Kotlin / MyBatis3 / �→ MyBatis3Simple 4 selection to choice. --> application server’s CPU cores and Numa resources cannot be fully utilized. For example, only one JVM is used to establish thousands of JDBC connections to TiDB. Example 1: TiDB is the bottleneck of user troubleshoot issues. 12.5.2.6.1 Troubleshooting tools Using the powerful troubleshooting tools of JVM is recommended when an issue occurs in your Java application and you do not know the application logic

targetRuntime: Used to specify the runtime target for generated code. It has MyBatis3DynamicSql / MyBatis3Kotlin / MyBatis3 / MyBatis3Simple 4 selection to �→ choice. --> TiDB 集群不在同一个地区，比如数据库的 DNS 均衡器和 TiDB 集群不在同一个地区。 • 客户端程序存在瓶颈，无法充分利用服务器的多 CPU 核或者多 Numa 资源，比如应用只使用一个 JVM 向 TiDB 建立上千个 JDBC 连接。 示例 1：用户响应时间的瓶颈在 TiDB 中 图 108: TiDB is the Bottleneck 在此 TPC-C 负载中： • 所有 应用发生问题并且不知道业务逻辑情况下，使用 JVM 强大的排查工具会比较有用。这里简单介绍几个 常用工具： jstack jstack 对应于 Go 中的 pprof/goroutine，可以比较方便地排查进程卡死的问题。 通过执行 jstack pid，即可输出目标进程中所有线程的线程 id 和堆栈信息。输出中默认只有 Java 堆栈，如果 希望同时输出 JVM 中的 C++ 堆栈，需要加 -m 选项。

application server’s CPU cores and Numa resources cannot be fully utilized. For example, only one JVM is used to establish thousands of JDBC connections to TiDB. In the Connection Count panel, you can troubleshoot issues. 12.5.2.6.1 Troubleshooting tools Using the powerful troubleshooting tools of JVM is recommended when an issue occurs in your Java application and you do not know the application logic target process. By default, only the Java stack is output. If you want to output the C++ stack in the JVM at the same time, add the -m option. By using jstack multiple times, you can easily locate the stuck

application server’s CPU cores and Numa resources cannot be fully utilized. For example, only one JVM is used to establish thousands of JDBC connections to TiDB. In the Connection Count panel, you can troubleshoot issues. 12.5.2.6.1 Troubleshooting tools Using the powerful troubleshooting tools of JVM is recommended when an issue occurs in your Java application and you do not know the application logic target process. By default, only the Java stack is output. If you want to output the C++ stack in the JVM at the same time, add the -m option. By using jstack multiple times, you can easily locate the stuck

application server’s CPU cores and Numa resources cannot be fully utilized. For example, only one JVM is used to establish thousands of JDBC connections to TiDB. In the Connection Count panel, you can troubleshoot issues. 12.5.2.6.1 Troubleshooting tools Using the powerful troubleshooting tools of JVM is recommended when an issue occurs in your Java application and you do not know the application logic target process. By default, only the Java stack is output. If you want to output the C++ stack in the JVM at the same time, add the -m option. By using jstack multiple times, you can easily locate the stuck

application server’s CPU cores and Numa resources cannot be fully utilized. For example, only one JVM is used to establish thousands of JDBC connections to TiDB. In the Connection Count panel, you can troubleshoot issues. 12.5.2.6.1 Troubleshooting tools Using the powerful troubleshooting tools of JVM is recommended when an issue occurs in your Java application and you do not know the application logic target process. By default, only the Java stack is output. If you want to output the C++ stack in the JVM at the same time, add the -m option. By using jstack multiple times, you can easily locate the stuck

TiDB 集群不在同一个地区，比如数据库的 DNS 均衡器和 TiDB 集群不在同一个地区。 • 客户端程序存在瓶颈，无法充分利用服务器的多 CPU 核或者多 Numa 资源，比如应用只使用一个 JVM 向 TiDB 建立上千个 JDBC 连接。 996 在 Connection Count （连接信息）面板中，你可以查看总的连接数和每个 TiDB 节点的连接数，并由此判断连接 总数是否正常，各 应用发生问题并且不知道业务逻辑情况下，使用 JVM 强大的排查工具会比较有用。这里简单介绍几个 常用工具： jstack jstack 对应于 Go 中的 pprof/goroutine，可以比较方便地排查进程卡死的问题。 通过执行 jstack pid，即可输出目标进程中所有线程的线程 id 和堆栈信息。输出中默认只有 Java 堆栈，如果 希望同时输出 JVM 中的 C++ 堆栈，需要加 -m 选项。 concurrency �→ 5 在 index scan 中获取 row keys 的最大线程数（每个 JVM 的所有任务共享）。 spark. �→ tispark. �→ table. �→ scan_ �→ concurrency �→ 512 在 table scan 中的最大线程数（每个 JVM 的所有任务共享）。 spark. �→ tispark. �→ request. �→ command

application server’s CPU cores and Numa resources cannot be fully utilized. For example, only one JVM is used to establish thousands of JDBC connections to TiDB. In the Connection Count panel, you can troubleshoot issues. 12.5.2.6.1 Troubleshooting tools Using the powerful troubleshooting tools of JVM is recommended when an issue occurs in your Java application and you do not know the application logic target process. By default, only the Java stack is output. If you want to output the C++ stack in the JVM at the same time, add the -m option. By using jstack multiple times, you can easily locate the stuck

TiDB 集群不在同一个地区，比如数据库的 DNS 均衡器和 TiDB 集群不在同一个地区。 • 客户端程序存在瓶颈，无法充分利用服务器的多 CPU 核或者多 Numa 资源，比如应用只使用一个 JVM 向 TiDB 建立上千个 JDBC 连接。 936 在 Connection Count （连接信息）面板中，你可以查看总的连接数和每个 TiDB 节点的连接数，并由此判断连接 总数是否正常，各 应用发生问题并且不知道业务逻辑情况下，使用 JVM 强大的排查工具会比较有用。这里简单介绍几个 常用工具： jstack jstack 对应于 Go 中的 pprof/goroutine，可以比较方便地排查进程卡死的问题。 通过执行 jstack pid，即可输出目标进程中所有线程的线程 id 和堆栈信息。输出中默认只有 Java 堆栈，如果 希望同时输出 JVM 中的 C++ 堆栈，需要加 -m 选项。 concurrency �→ 5 在 index scan 中获取 row keys 的最大线程数（每个 JVM 的所有任务共享）。 spark. �→ tispark. �→ table. �→ scan_ �→ concurrency �→ 512 在 table scan 中的最大线程数（每个 JVM 的所有任务共享）。 spark. �→ tispark. �→ request. �→ command

application server’s CPU cores and Numa resources cannot be fully utilized. For example, only one JVM is used to establish thousands of JDBC connections to TiDB. In the Connection Count panel, you can troubleshoot issues. 12.5.2.6.1 Troubleshooting tools Using the powerful troubleshooting tools of JVM is recommended when an issue occurs in your Java application and you do not know the application logic target process. By default, only the Java stack is output. If you want to output the C++ stack in the JVM at the same time, add the -m option. By using jstack multiple times, you can easily locate the stuck