Data types - ApsaraDB for HBase - Alibaba Cloud Documentation Center

Basic data types


Data type	Supported constant	Description
bigint	integer	The 64-bit signed long value.
blob	blob	A number of bytes. The length of these bytes ranges from 0 to infinity.
boolean	boolean	Valid values: true and false.
counter	integer	The counter column that contains 64-bit signed values.
date	integer, string	The date.
decimal	integer, float	The variable-granularity decimal value.
double	integer, float	The 64-bit IEEE-754 floating value.
float	integer, float	The 32-bit IEEE-754 floating value.
inet	string	The IP address. Valid types: IPv4 and IPv6. All inputs are stored as strings because no IP data type exists.
int	integer	32-bit signed int
smallint	integer	16-bit signed int
text	string	The string in the UTF-8 format.
time	integer, string	The time that is measured in nanoseconds.
timestamp	integer, string	The timestamp that is measured in microseconds.
timeuuid	uuid	The Version 1 UUID. In most cases, the UUID is used as a conflict-free timestamp.
tinyint	integer	8-bit signed int
uuid	uuid	The UUID.
varint	integer	The integer of a variable granularity.
nested	string	The JSON string that can be nested.

Counter

The counter type that is used to specify a counter column. The counter type of Lindorm CQL is slightly different from the counter type of Cassandra. The counter tables of Cassandra contain only counter columns. However, the counter tables of Lindorm CQL can contain counter columns and non-counter columns. Example:

CREATE TABLE persioninfo  (name text primary key, age counter, address text);
UPDATE persioninfo SET age = age + 2 WHERE name = 'my';
INSERT INTO persioninfo  (name, age, address) VALUES ( 'my', 30, 'hz');

Timestamp

You can specify an integer or string to represent a point in time. The following example is used to represent the 04:05:00 March 2, 2011 GMT point in time.

CREATE TABLE persioninfo ( name text PRIMARY KEY , birthtime timestamp , address text );
INSERT INTO persioninfo (name, birthtime, address) VALUES ( 'my', 1299038700000, 'hz');
INSERT INTO persioninfo (name, birthtime, address) VALUES ( 'mm', '2011-03-02 04:05+0000', 'hz');
INSERT INTO persioninfo (name, birthtime, address) VALUES ( 'lucy', '2011-03-02 04:05:00+0000', 'bj');

Time

You can use an integer or string to represent a time. If a string is used to represent a time, the string must be specified in the hh:mm:ss[.fffffffff] format. The [.fffffffff] parameter is optional. Example:

CREATE TABLE persioninfo ( name text PRIMARY KEY , birthtime time , adress text );
INSERT INTO persioninfo (name,birthtime,adress) VALUES ( 'lili', '08:12:54', 'hz');

date

You can use an integer or string to represent a date. If a string is used to represent a time, the string must be specified in the yyyy-mm-dd format. Example:

 CREATE TABLE persioninfo ( name text PRIMARY KEY , birthday date , address text );
 INSERT INTO persioninfo (name,birthday,address) VALUES ( 'lucy', '2021-02-01', 'beijing');

nested

The nested type. When you create a table, you can set the data type of a column to nested. You can write JSON data as strings to the column. You can use a SELECT statement in which the search_query clause is specified to run a nested sub-query on the column. To implement this sub-query, the search index feature of Lindorm must be used in combination with the preceding setting. Note: If you execute a DESCRIBE statement on the table that contains the column, the statement still returns the text data for the column. In addition, nested data is still displayed as text data on the open source driver side. Example:

 CREATE TABLE persioninfo ( name text PRIMARY KEY , info nested , address text );
 CREATE SEARCH INDEX  sidx ON ctable3 WITH COLUMNS (info ,address );
 INSERT INTO ctable3 (name,info,address) VALUES ( 'lilei', '{"school":"alibaba"}', 'hz');
 SELECT * FROM ctable3 WHERE search_query = '{+address:hz}';

Unsupported data types

Collection data types: SET, LIST, and MAP.
User-defined types.
Basic data types: Duration.