TiDB 4.0 课程 21 天定制学习计划_day4


2.1 When to use TiDB platform
2.2 How-to-connect-TiDB-platform






When to Use TiDB

OLTP-like workload

when you need Random,Real-time Read/Write access to your big data(billions of rows),with
  • ACID compliance

  • Secondary index support

  • MySQL syntax

    Real-time HTAP

  • Real-time HTAP(Hybrid transactional/analytical processing)

  • When you have OLTP-like workload using TiDB,and you want to perform OLAP query in place with the help of TiFlash
    a,with fresh data
    b,with zero interfarence on OLTP performance

  • Data integration
    a,when you have multiple data sources like(OLTP databases,streaming ingestion and etc),you want to perform OLAT query on the integrated data set.

Connect with Spark Eco-system via TiSpark

With TiSpark you can use Spark to process data within TiDB platform inplace without moving your data out.

It's useful when you want workload like:
  • Iterative processing
    a,like traditional ETL job
  • Joining datasets
    a,joining large datasets from multiple data sources
    b,a lot of shuffling and sorting is needed

TiDB may not be your best choice when

  • Your data can fit on a single server
  • You need to perform heavy analytics tasks
    a,scanning and aggregation on large data sets,that intermediate results can not fit in single server’s memory
  • You need sub-millisecond latency
    a,take a look at Reads?

TiDB speaks MySQL protocol

TiDB has

  • (almost)same syntax
  • same write protocol

as Mysql 5.7


  • CLI
  • GUI


  • Python
  • Java
  • Golang

ORM(Object-Relational Mapping)

  • Java
  • Python
  • Go(Gorm,xorm)