2. $ cd /opt/spark/data/tispark-sample-data 3. $ mysql -h tidb -P 4000 -u root < dss.ddl 当样本数据加载到 TiDB 集群之后，可以使用 docker-compose exec tispark-master /opt/spark/bin/spark-shell 来访 问 Spark shell。 1. $ docker-compose exec tispark-master /opt/spark/bin/spark-shell 2. ... 3. Spark context available as 'sc' (master = local[*], app id = local-1527045927617). 4. Spark session available as 'spark'. 5. Welcome to 6. 16. scala> import org.apache.spark.sql.TiContext 17. ... 18. scala> val ti = new TiContext(spark) 19. ... 20. scala> ti.tidbMapDatabase("TPCH_001") 21. ... 22. scala> spark.sql("select count(*) from lineitem")

XGBoost Spark SparkApplication Spark Operator 在线应用 无状态服务 Node Node Node Node Node Node 无状态服务 无状态服务 无状态服务 无状态服务 无状态服务 无状态服务 TF PS 1 TF Worker 1 TF Worker 2 spark driver 1 spark exec 1 spark exec 212/19

nodes based on the actual status of the system. • TiSpark • If your data needs to be analyzed with Spark, deploy TiSpark. For specific process, see TiSpark User Guide. 142 3.4.4 Data preparation After topology. TiSpark is a component built for running Apache Spark on top of TiDB/TiKV to answer complex OLAP queries. It brings benefits of both the Spark platform and the distributed TiKV cluster to TiDB and deployed in the home directory of the user. 5.3.6.3 Prerequisites TiSpark is based on the Apache Spark cluster, so before you start the TiDB cluster that contains TiSpark, you must ensure that Java Runtime

nodes based on the actual status of the system. • TiSpark • If your data needs to be analyzed with Spark, deploy TiSpark. For specific process, see TiSpark User Guide. 112 3.4.4 Data preparation After topology. TiSpark is a component built for running Apache Spark on top of TiDB/TiKV to answer complex OLAP queries. It brings benefits of both the Spark platform and the distributed TiKV cluster to TiDB and deployed in the home directory of the user. 5.3.7.3 Prerequisites TiSpark is based on the Apache Spark cluster, so before you start the TiDB cluster that contains TiSpark, you must ensure that Java Runtime

nodes based on the actual status of the system. • TiSpark • If your data needs to be analyzed with Spark, deploy TiSpark. For specific process, see TiSpark User Guide. 115 3.4.4 Data preparation After topology. TiSpark is a component built for running Apache Spark on top of TiDB/TiKV to answer complex OLAP queries. It brings benefits of both the Spark platform and the distributed TiKV cluster to TiDB and deployed in the home directory of the user. 5.3.7.3 Prerequisites TiSpark is based on the Apache Spark cluster, so before you start the TiDB cluster that contains TiSpark, you must ensure that Java Runtime

nodes based on the actual status of the system. • TiSpark • If your data needs to be analyzed with Spark, deploy TiSpark. For specific process, see TiSpark User Guide. 137 3.4.4 Data preparation After topology. TiSpark is a component built for running Apache Spark on top of TiDB/TiKV to answer complex OLAP queries. It brings benefits of both the Spark platform and the distributed TiKV cluster to TiDB and deployed in the home directory of the user. 5.3.6.3 Prerequisites TiSpark is based on the Apache Spark cluster, so before you start the TiDB cluster that contains TiSpark, you must ensure that Java Runtime

nodes based on the actual status of the system. • TiSpark • If your data needs to be analyzed with Spark, deploy TiSpark. For specific process, see TiSpark User Guide. 112 3.4.4 Data preparation After topology. TiSpark is a component built for running Apache Spark on top of TiDB/TiKV to answer complex OLAP queries. It brings benefits of both the Spark platform and the distributed TiKV cluster to TiDB and deployed in the home directory of the user. 5.3.6.3 Prerequisites TiSpark is based on the Apache Spark cluster, so before you start the TiDB cluster that contains TiSpark, you must ensure that Java Runtime

会成为主要瓶颈，也容易产生读写热点。此时 TiFlash 节点数与 OLAP 计算量有较复杂非线性关系，需要根据具体系统状态调整节点数量。 97 • TiSpark – 如果你的业务需要基于 Spark 进行分析，请部署 TiSpark。具体步骤，请参阅TiSpark 用户指南。 3.4.4 HTAP 数据准备 TiFlash 部署完成后并不会自动同步数据，你需要指定需要同步到 TiFlash TiSpark 部署拓扑 本文介绍 TiSpark 部署的拓扑，以及如何在最小拓扑的基础上同时部署 TiSpark。TiSpark 是 PingCAP 为解决用户复 杂 OLAP 需求而推出的产品。它借助 Spark 平台，同时融合 TiKV 分布式集群的优势，和 TiDB 一起为用户一站式 解决 HTAP (Hybrid Transactional/Analytical Processing) 的需求。 关于 tispark_masters: - host: 10.0.1.21 ### NOTE: multiple worker nodes on the same host is not supported by Spark tispark_workers: - host: 10.0.1.22 - host: 10.0.1.23 monitoring_servers: - host: 10.0.1.10 grafana_servers:

会成为主要瓶颈，也容易产生读写热点。此时 TiFlash 节点数与 OLAP 计算量有较复杂非线性关系，需要根据具体系统状态调整节点数量。 81 • TiSpark – 如果你的业务需要基于 Spark 进行分析，请部署 TiSpark。具体步骤，请参阅TiSpark 用户指南。 3.4.4 HTAP 数据准备 TiFlash 部署完成后并不会自动同步数据，你需要指定需要同步到 TiFlash TiSpark 部署拓扑 本文介绍 TiSpark 部署的拓扑，以及如何在最小拓扑的基础上同时部署 TiSpark。TiSpark 是 PingCAP 为解决用户复 杂 OLAP 需求而推出的产品。它借助 Spark 平台，同时融合 TiKV 分布式集群的优势，和 TiDB 一起为用户一站式 解决 HTAP (Hybrid Transactional/Analytical Processing) 的需求。 关于 tispark_masters: - host: 10.0.1.21 ### NOTE: multiple worker nodes on the same host is not supported by Spark tispark_workers: - host: 10.0.1.22 - host: 10.0.1.23 monitoring_servers: - host: 10.0.1.10 grafana_servers:

I/O 会成为主要瓶颈，也容易产生读写热点。此时 TiFlash 节点数与 OLAP 计算量有较复杂非线性关系，需要根据具体系统状态调整节点数量。 • TiSpark – 如果你的业务需要基于 Spark 进行分析，请部署 TiSpark。具体步骤，请参阅TiSpark 用户指南。 3.4.4 HTAP 数据准备 TiFlash 部署完成后并不会自动同步数据，你需要指定需要同步到 TiFlash TiSpark 部署拓扑 本文介绍 TiSpark 部署的拓扑，以及如何在最小拓扑的基础上同时部署 TiSpark。TiSpark 是 PingCAP 为解决用户复 杂 OLAP 需求而推出的产品。它借助 Spark 平台，同时融合 TiKV 分布式集群的优势，和 TiDB 一起为用户一站式 解决 HTAP (Hybrid Transactional/Analytical Processing) 的需求。 关于 tispark_masters: - host: 10.0.1.21 ### NOTE: multiple worker nodes on the same host is not supported by Spark tispark_workers: - host: 10.0.1.22 - host: 10.0.1.23 monitoring_servers: - host: 10.0.1.10 grafana_servers: