All Products
Document Center


Last Updated:Jan 31, 2024

In MaxCompute, tables are used to store data. Logically, a table is a two-dimensional structure that consists of rows and columns. Each row represents a record, and each column represents a field whose values are of the same data type. Similar to tables in traditional relational databases, MaxCompute tables can store structured data and use SQL to query and analyze data.

Tables are the input and output objects of all computing tasks in MaxCompute. You can create a table, delete a table, and import data to a table. For more information, see Table operations.


You can manage tables on the DataStudio page of DataWorks. For example, you can create MaxCompute tables, add MaxCompute tables to your favorites, manage data lifecycles, modify table schemas, and manage permissions on tables, resources, or functions.

Internal tables and external tables are supported by MaxCompute. MaxCompute V2.0 and later support external tables.

  • Data of internal tables is stored in MaxCompute. Columns in internal tables can be of any data types that are supported by MaxCompute. For more information, see Data type editions.

  • Data of external tables is not stored in MaxCompute. Instead, the data can be stored in Object Storage Service (OSS) or Tablestore. MaxCompute records only metadata of external tables. You can use external tables of MaxCompute to process unstructured data that is stored in OSS or Tablestore. The unstructured data includes video, audio, genetic, meteorological, or geographic data.


You can create an external table, read data from an external table, or write data to an external table. For more information about external tables, see Overview.