CREATE TABLE

Purpose

This statement creates a new table in the database.

Syntax

CREATE [hint_options] [TEMPORARY] TABLE [IF NOT EXISTS] table_name
      (table_definition_list) [table_option_list] [partition_option] [[MERGE_ENGINE = {delete_insert | partial_update}] table_column_group_option] [IGNORE | REPLACE] [AS] select;

CREATE TABLE [TEMPORARY] [IF NOT EXISTS] table_name
      LIKE table_name;

table_definition_list:
    table_definition [, table_definition ...]

table_definition:
      column_definition_list
    | [CONSTRAINT [constraint_name]] PRIMARY KEY index_desc
    | [CONSTRAINT [constraint_name]] UNIQUE {INDEX | KEY}
            [index_name] index_desc
    | [CONSTRAINT [constraint_name]] FOREIGN KEY
            [index_name] index_desc
            REFERENCES reference_definition
            [match_action][opt_reference_option_list]
    | [FULLTEXT] {INDEX | KEY} [index_name] [index_type] (key_part,...) [WITH PARSER tokenizer_option] [PARSER_PROPERTIES[=](parser_properties_list)]
      [index_option_list] [index_column_group_option]
    | index_json_clause
    | [CONSTRAINT [constraint_name]] CHECK(expression) constranit_state

column_definition_list:
    column_definition [, column_definition ...]

column_definition:
     column_name data_type
         [DEFAULT const_value] [AUTO_INCREMENT]
         [NULL | NOT NULL] [[PRIMARY] KEY] [UNIQUE [KEY]] [COMMENT string_value] [SKIP_INDEX(skip_index_option_list)]
   | column_name data_type
         [GENERATED ALWAYS] AS (expr) [VIRTUAL | STORED]
         [opt_generated_column_attribute]

skip_index_option_list:
    skip_index_option [,skip_index_option ...]

skip_index_option:
    MIN_MAX
    | SUM

index_desc:
   (column_desc_list) [index_type] [index_option_list]

match_action:
   MATCH {SIMPLE | FULL | PARTIAL}

opt_reference_option_list:
   reference_option [,reference_option ...]

reference_option:
   ON {DELETE | UPDATE} {RESTRICT | CASCADE | SET NULL | NO ACTION | SET DEFAULT}

tokenizer_option:
    SPACE
    | NGRAM
    | BENG
    | IK
    | NGRAM2

parser_properties_list:
    parser_properties, [parser_properties]

parser_properties:
    min_token_size = int_value
    | max_token_size = int_value
    | ngram_token_size = int_value
    | ik_mode = 'char_value'
    | min_ngram_size = int_value
    | max_ngram_size = int_value

key_part:
    {index_col_name [(length)] | (expr)} [ASC | DESC]

index_type:
    USING BTREE

index_option_list:
    index_option [ index_option ...]

index_option:
      [GLOBAL | LOCAL]
    | block_size
    | compression
    | STORING(column_name_list)
    | COMMENT string_value
    | STORAGE_CACHE_POLICY(storage_cache_policy_option)

table_option_list:
    table_option [ table_option ...]

table_option:
      [DEFAULT] {CHARSET | CHARACTER SET} [=] charset_name
    | [DEFAULT] COLLATE [=] collation_name
    | table_tablegroup
    | block_size
    | lob_inrow_threshold [=] num
    | compression
    | AUTO_INCREMENT [=] INT_VALUE
    | COMMENT string_value
    | ROW_FORMAT [=] REDUNDANT|COMPACT|DYNAMIC|COMPRESSED|DEFAULT
    | PCTFREE [=] num
    | parallel_clause
    | DUPLICATE_SCOPE [=] 'none|cluster'
    | TABLE_MODE [=] 'table_mode_value'
    | auto_increment_cache_size [=] INT_VALUE
    | READ {ONLY | WRITE}
    | ORGANIZATION [=] {INDEX | HEAP}
    | enable_macro_block_bloom_filter [=] {True | False}
    | DYNAMIC_PARTITION_POLICY [=] (dynamic_partition_policy_list)
    | SEMISTRUCT_ENCODING_TYPE [=] 'encoding' # Deprecated since V4.4.1. Use SEMISTRUCT_PROPERTIES instead.
    | MICRO_BLOCK_FORMAT_VERSION [=] {1|2}
    | STORAGE_CACHE_POLICY (storage_cache_policy_option)

parallel_clause:
    {NOPARALLEL | PARALLEL integer}

table_mode_value:
    NORMAL
    | QUEUING
    | MODERATE
    | SUPER
    | EXTREME

dynamic_partition_policy_list:
    dynamic_partition_policy_option [, dynamic_partition_policy_option ...]

dynamic_partition_policy_option:
    ENABLE = {true | false}
    | TIME_UNIT = {'hour' | 'day' | 'week' | 'month' | 'year'}
    | PRECREATE_TIME = {'-1' | '0' | 'n {hour | day | week | month | year}'}
    | EXPIRE_TIME = {'-1' | '0' | 'n {hour | day | week | month | year}'}
    | TIME_ZONE = {'default' | 'time_zone'}
    | BIGINT_PRECISION = {'none' | 'us' | 'ms' | 's'}

partition_option:
      PARTITION BY HASH(expression)
      [subpartition_option] PARTITIONS partition_count
    | PARTITION BY KEY([column_name_list])
      [subpartition_option] PARTITIONS partition_count
    | PARTITION BY RANGE {(expression) | COLUMNS (column_name_list)}
      [subpartition_option] (range_partition_list) [STORAGE_CACHE_POLICY = {"hot" | "auto" | "none"}]
    | PARTITION BY LIST {(expression) | COLUMNS (column_name_list)}
      [subpartition_option] PARTITIONS partition_count
    | PARTITION BY RANGE [COLUMNS]([column_name_list]) [SIZE('size_value')] (range_partition_list)

subpartition_option:
      SUBPARTITION BY HASH(expression)
      SUBPARTITIONS subpartition_count
    | SUBPARTITION BY KEY(column_name_list)
      SUBPARTITIONS subpartition_count
    | SUBPARTITION BY RANGE {(expression) | COLUMNS (column_name_list)}
      (range_subpartition_list) [STORAGE_CACHE_POLICY = {"hot" | "auto" | "none"}]
    | SUBPARTITION BY LIST(expression)

storage_cache_policy_option:
    GLOBAL = {"hot" | "auto" | "none"}
    | timeline_strategy_list

timeline_strategy_list:
    BOUNDARY_COLUMN = column_name [,BOUNDARY_COLUMN_UNIT = {"s"| "ms"}] ,HOT_RETENTON = intnum retention_time_unit

retention_time_unit:
    YEAR
    | MONTH
    | WEEK
    | DAY
    | HOUR
    | MINUTE
range_partition_list:
    range_partition [, range_partition ...]

range_partition:
    PARTITION partition_name
    VALUES LESS THAN {(expression_list) | MAXVALUE}
    [STORAGE_CACHE_POLICY = {"hot" | "auto" | "none"}]

range_subpartition_list:
    range_subpartition [, range_subpartition ...]

range_subpartition:
    SUBPARTITION subpartition_name
    VALUES LESS THAN {(expression_list) | MAXVALUE}
    [STORAGE_CACHE_POLICY = {"hot" | "auto" | "none"}]

expression_list:
    expression [, expression ...]

column_name_list:
    column_name [, column_name ...]

partition_name_list:
    partition_name [, partition_name ...]

partition_count | subpartition_count:
    INT_VALUE

table_column_group_option/index_column_group_option:
      WITH COLUMN GROUP(all columns)
    | WITH COLUMN GROUP(each column)
    | WITH COLUMN GROUP(all columns, each column)

index_json_clause:
    [UNIQUE] INDEX idx_json_name((CAST(json_column_name->'$.json_field_name' AS UNSIGNED ARRAY)))
    | INDEX idx_json_name(column_name, [column_name, ...] (CAST(json_column_name->'$.json_field_name' AS CHAR(n) ARRAY)))

Parameters

table_mode_value

• NORMAL: The default value, indicating normal mode. In this mode, the probability of triggering a major compaction after a data dump is very low.
• QUEUING: In this mode, the probability of triggering a major compaction after a data dump is low.
• MODERATE: Indicates a moderate mode. In this mode, the probability of triggering a major compaction after a data dump is moderate.
• SUPER: Indicates a super mode. In this mode, the probability of triggering a major compaction after a data dump is high.
• EXTREME: Indicates an extreme mode. In this mode, the probability of triggering a major compaction after a data dump is high.

For more information about major compactions, see Adaptive compaction.

tokenizer_option

• SPACE: The default value, indicating that the tokenizer splits text by spaces. You can specify the following properties:

  Property	Value Range
  min_token_size	[1, 16]
  max_token_size	[10, 84]

• NGRAM: Indicates the N-Gram tokenizer for Chinese text. You can specify the following properties:

  Property	Value Range
  ngram_token_size	[1, 10]

• NGRAM2: Indicates the tokenizer that splits text into continuous characters of length between min_ngram_size and max_ngram_size. You can specify the following properties:

  Property	Value Range
  min_ngram_size	[1, 16]
  max_ngram_size	[1, 16]

• BENG: Indicates the Beng tokenizer for basic English text. You can specify the following properties:

  Property	Value Range
  min_token_size	[1, 16]
  max_token_size	[10, 84]

• IK: Indicates the IK tokenizer for Chinese text. Currently, only the utf-8 character set is supported. You can specify the following property:

  Property	Value Range
  ik_mode	smart max_word

parser_properties

• min_token_size: The minimum token length. The default value is 3, and the value ranges from 1 to 16.

• max_token_size: The maximum token length. The default value is 84, and the value ranges from 10 to 84.

• ngram_token_size: The token length for the NGRAM tokenizer. This property is valid only for the NGRAM tokenizer. The default value is 2, and the value ranges from 1 to 10.

• ik_mode: The tokenization mode for the IK tokenizer. Valid values:

  • smart: The default value. The tokenizer uses the dictionary to improve tokenization accuracy. The boundaries of the dictionary words are prioritized, which may reduce unnecessary expansions.
  • max_word: The tokenizer identifies the words defined in the dictionary but does not affect the maximum expansion of tokenization. Even if the dictionary defines the words, the max_word mode still tries to split the text into more words.

• min_ngram_size: The minimum token length. The value ranges from 1 to 16.

• max_ngram_size: The maximum token length. The value ranges from 1 to 16.

Here is an example:

CREATE TABLE tbl1 (col1 VARCHAR(200), col2 TEXT,
    FULLTEXT INDEX fidx(col2) WITH PARSER ik PARSER_PROPERTIES=(ik_mode='max_word'));

dynamic_partition_policy_option

• ENABLE = {true | false}: Optional. Indicates whether to enable dynamic partition management. This parameter can be modified. Valid values:

  • true: The default value. Indicates to enable dynamic partition management.
  • false: Indicates to disable dynamic partition management.

• TIME_UNIT = {'hour' | 'day' | 'week' | 'month' | 'year'}: Required. Indicates the time unit for partitioning, that is, the interval for automatically creating partition boundaries. This parameter cannot be modified. Valid values:

  • hour: Partitions are created by hour.
  • day: Partitions are created by day.
  • week: Partitions are created by week.
  • month: Partitions are created by month.
  • year: Partitions are created by year.

• PRECREATE_TIME = {'-1' | '0' | 'n {hour | day | week | month | year}'}: Optional. Indicates the precreation time. When dynamic partition management is scheduled, partitions are precreated so that maximum partition upper bound > now() + precreate_time. This parameter can be modified. Valid values:

  • -1: The default value. Indicates not to precreate partitions.
  • 0: Indicates to precreate only the current partition.
  • n {hour | day | week | month | year}: Indicates the time span for precreation. For example, 3 hour indicates to precreate partitions for the last 3 hours.

  Note

  • When multiple partitions are to be precreated, the interval between partition boundaries is TIME_UNIT.
  • The boundary of the first precreated partition is the ceiling of the maximum existing partition boundary based on TIME_UNIT.

• EXPIRE_TIME = {'-1' | '0' | 'n {hour | day | week | month | year}'}: Optional. Indicates the partition expiration time. When dynamic partition management is scheduled, all partitions with upper bounds < now() - expire_time are deleted. This parameter can be modified. Valid values:

  • -1: The default value. Indicates that partitions do not expire.
  • 0: Indicates that all partitions except the current one expire.
  • n {hour | day | week | month | year}: Indicates the partition expiration time. For example, 1 day indicates that partitions expire after 1 day.

• TIME_ZONE = {'default' | 'time_zone'}: Optional. Indicates the time zone for determining the current time and the size of the partitioning key of the date, datetime, and year types. This parameter cannot be modified. Valid values:

  • default: The default value. Indicates not to configure a time zone and to use the tenant time zone. For types other than date, datetime, and year, the time_zone field must be default.
  • time_zone: Indicates a custom time zone offset. For example, +8:00 is a time zone offset.

• BIGINT_PRECISION = {'none'| 'ms' | 's'}: Optional. Indicates the timestamp precision of the bigint type partitioning key. This parameter cannot be modified. Valid values:

  • none: The default value. Indicates no precision (the partitioning key is not of the bigint type).
  • ms: Indicates millisecond precision.
  • s: Indicates second precision.

For more information about creating a dynamic partition table, see Create a dynamic partition table.

Here is an example:

CREATE TABLE tbl2 (col1 INT, col2 DATETIME)
    DYNAMIC_PARTITION_POLICY(
        ENABLE = true,
        TIME_UNIT = 'hour',
        PRECREATE_TIME = '3 hour',
        EXPIRE_TIME = '1 day',
        TIME_ZONE = '+8:00',
        BIGINT_PRECISION = 'none')
    PARTITION BY RANGE COLUMNS (col2)(
        PARTITION P0 VALUES LESS THAN ('2024-11-11 13:30:00')
        );

storage_cache_policy

Create a table-level storage_cache_policy

The STORAGE_CACHE_POLICY option is optional. If you do not specify this option, the default value of the tenant-level configuration parameter default_storage_cache_policy is used. The attributes in the storage_cache_policy_option parameter are defined in the Key-Value format. The semantics of each attribute are as follows:

• GLOBAL = {"hot"|"auto"|"none"}: specifies the hot cache strategy for the full table data. Valid values:

  • hot: specifies that all data in the table is hot data. If the cache disk space is sufficient, all data in the table will be cached to the local cache disk.

  • auto: specifies that the hot data in the table is automatically identified by the system.

  • none: specifies that the cache strategy for the index follows the value of STORAGE_CACHE_POLICY of the parent table.

    Notice

    The none attribute can only be used for indexes.

• timeline_strategy_list: specifies the list of time-axis strategy parameters. Parameters are separated by commas. The hot cache time-axis strategy supports a mechanism to determine whether partition data is hot data based on time. The system automatically adjusts the partition data on the local cache disk based on the configured strategy.

  Note

  A time-axis strategy specifies that hot data is cached based on the partition range defined by the Range partition. When a certain time condition is met, the data in the partition is determined to be hot data. When you use a time-axis strategy, note the following:

  • Only Range-partitioned tables can use a time-axis strategy. A Range partition can be a primary or secondary partition, but a double Range partition cannot be used. This is because it is impossible to determine which Range partition to use for the time-axis strategy.
  • When you use a time-axis strategy, the partition expression can contain only column names and cannot contain expressions. For example, PARTITION BY RANGE COLUMNS(expr(col3)) is invalid.
  • BOUNDARY_COLUMN must be a partitioning key. If the partitioning key contains multiple columns, BOUNDARY_COLUMN must be the first column to determine whether the partition has expired.

• BOUNDARY_COLUMN = column_name: specifies the column to determine hot data. The column can be an integer column of the BIGINT or INT type, in the Unix timestamp format, or a time column of the TIMESTAMP, DATE, DATETIME, or YEAR type.

  • If the BOUNDARY_COLUMN is an integer column, the table can be partitioned by Range or Range Columns.

  • If the BOUNDARY_COLUMN is a time column, the table can be partitioned only by Range Columns.

  • BOUNDARY_COLUMN_UNIT = {"s" | "ms"}: specifies the time unit of the BOUNDARY_COLUMN parameter. This parameter can be set only when the BOUNDARY_COLUMN is an integer column. When an integer value is used as a timestamp, you must specify the timestamp unit to avoid incorrect parsing of the timestamp. Valid values:

    • If the partition column is of the INT type, the BOUNDARY_COLUMN_UNIT can be only s.

    • If the partition column is of the BIGINT type, the BOUNDARY_COLUMN_UNIT can be s or ms.

      Notice

      If the format is not a Unix timestamp, the system cannot correctly identify the time represented by the INT type.

• HOT_RETENTION = intnum retention_time_unit: specifies the time range for hot data.

  • intnum: specifies an integer.

  • retention_time_unit: specifies the time unit. Valid values:

    • YEAR: specifies years.
    • MONTH: specifies months.
    • WEEK: specifies weeks.
    • DAY: specifies days.
    • HOUR: specifies hours.
    • MINUTE: specifies minutes.

Create a partition-level storage_cache_policy

The STORAGE_CACHE_POLICY option is optional for partitions. Valid values:

• hot: specifies that all data in the partition is hot data. If the cache disk space is sufficient, all data in the partition will be cached to the local cache disk.
• auto: specifies that the hot data in the partition is automatically identified by the system.
• none: specifies that the hot cache strategy for the partition follows the value of STORAGE_CACHE_POLICY of the parent partition. This is the default value.

Examples

• Create a database table.

  obclient> CREATE TABLE tbl1 (c1 INT PRIMARY KEY, c2 VARCHAR(50));
  Query OK, 0 rows affected
  

• Create a table with an index.

  obclient> CREATE TABLE tbl2 (c1 INT PRIMARY KEY, c2 INT, c3 INT, INDEX i1 (c2));
  Query OK, 0 rows affected
  

• Create a table tbl3 with a function-based index.

  obclient> CREATE TABLE tbl3 (c1 INT, c2 INT, INDEX i1 ((c1+1)), UNIQUE KEY ((c1+c2)));
  Query OK, 0 rows affected
  
  obclient> SHOW CREATE TABLE tbl3;
  +-------+----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
  | Table | Create Table                                                                                                                                                                                                                                                                                                                                                                     |
  +-------+----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
  | tbl3    | CREATE TABLE `tbl3` (
  `c1` int(11) DEFAULT NULL,
  `c2` int(11) DEFAULT NULL,
  UNIQUE KEY `functional_index` ((`c1` + `c2`)) BLOCK_SIZE 16384 LOCAL,
  KEY `i1` ((`c1` + 1)) BLOCK_SIZE 16384 LOCAL
  ) DEFAULT CHARSET = utf8mb4 ROW_FORMAT = DYNAMIC COMPRESSION = 'zstd_1.3.8' REPLICA_NUM = 2 BLOCK_SIZE = 16384 USE_BLOOM_FILTER = FALSE TABLET_SIZE = 134217728 PCTFREE = 0 |
  +-------+----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
  

• Create a table with 8 hash partitions.

  obclient> CREATE TABLE tbl4 (c1 INT PRIMARY KEY, c2 INT) PARTITION BY HASH(c1) PARTITIONS 8;
  Query OK, 0 rows affected
  

• Create a table with range partitions as primary partitions and key partitions as subpartitions.

  obclient> CREATE TABLE tbl5 (c1 INT, c2 INT, c3 INT) PARTITION BY RANGE(c1)
       SUBPARTITION BY KEY(c2, c3) SUBPARTITIONS 5
       (PARTITION p0 VALUES LESS THAN(0), PARTITION p1 VALUES LESS THAN(100));
  Query OK, 0 rows affected
  

• Create a table with one column of type gbk and another column of type utf8.

  obclient> CREATE TABLE tbl6 (c1 VARCHAR(10),
                   c2 VARCHAR(10) CHARSET GBK COLLATE gbk_bin)
               DEFAULT CHARSET utf8 COLLATE utf8mb4_general_ci;
  Query OK, 0 rows affected
  

• Enable encoding and use zstd compression. Set the space reserved for macroblocks to 5%.

  obclient> CREATE TABLE tbl7 (c1 INT, c2 INT, c3 VARCHAR(64))
        COMPRESSION 'zstd_1.0'
        ROW_FORMAT DYNAMIC
        PCTFREE 5;
  Query OK, 0 rows affected
  

• Create a table tbl8 and set the parallelism to 3.

  obclient> CREATE TABLE tbl8(c1 INT PRIMARY KEY, c2 INT) PARALLEL 3;
  Query OK, 0 rows affected
  

• Use an auto-increment column as the partitioning key.

  obclient> CREATE TABLE tbl9(inv_id BIGINT NOT NULL AUTO_INCREMENT,c1 BIGINT,
  PRIMARY KEY (inv_id) ) PARTITION BY HASH(inv_id) PARTITIONS 8;
  Query OK, 0 rows affected
  

• Specify a foreign key for the ref_t2 table. When an UPDATE operation affects the key values in the parent table that match the rows in the child table, execute the SET NULL operation.

  obclient> CREATE TABLE ref_t1(c1 INT PRIMARY KEY, C2 INT);
  Query OK, 0 rows affected
  
  obclient> CREATE TABLE ref_t2(c1 INT PRIMARY KEY, C2 INT,FOREIGN KEY(c2) REFERENCES ref_t1(c1) ON UPDATE SET NULL);
  Query OK, 0 rows affected
  

• Create a table tbl10 with a CHECK constraint and view the constraint information.

  obclient> CREATE TABLE tbl10 (col1 INT, col2 INT, col3 INT, CONSTRAINT equal_check1 CHECK(col1 = col3 * 2));
  Query OK, 0 rows affected
  
  obclient> SHOW CREATE TABLE tbl10;
  +-------+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
  | Table | Create Table                                                                                                                                                                                                                                                                                                                                               |
  +-------+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
  | tbl10  | CREATE TABLE `tbl10` (
    `col1` int(11) DEFAULT NULL,
    `col2` int(11) DEFAULT NULL,
    `col3` int(11) DEFAULT NULL,
    CONSTRAINT `equal_check1` CHECK ((`col1` = (`col3` * 2)))
  ) DEFAULT CHARSET = utf8mb4 ROW_FORMAT = DYNAMIC COMPRESSION = 'zstd_1.3.8' REPLICA_NUM = 1 BLOCK_SIZE = 16384 USE_BLOOM_FILTER = FALSE TABLET_SIZE = 134217728 PCTFREE = 0 |
  +-------+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
  1 row in set
  

• Create a table tbl11 based on the tbl10 table and view the table information.

  obclient> CREATE TABLE tbl11 LIKE tbl10;
  Query OK, 0 rows affected
  
  obclient> SHOW CREATE TABLE tbl11;
  +-------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
  | Table | Create Table                                                                                                                                                                                                                                                                                                                                                                |
  +-------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
  | tbl11  | CREATE TABLE `tbl11` (
    `col1` int(11) DEFAULT NULL,
    `col2` int(11) DEFAULT NULL,
    `col3` int(11) DEFAULT NULL,
    CONSTRAINT `tbl11_OBCHECK_1650793233327894` CHECK ((`col1` = (`col3` * 2)))
  ) DEFAULT CHARSET = utf8mb4 ROW_FORMAT = DYNAMIC COMPRESSION = 'zstd_1.3.8' REPLICA_NUM = 1 BLOCK_SIZE = 16384 USE_BLOOM_FILTER = FALSE TABLET_SIZE = 134217728 PCTFREE = 0 |
  +-------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
  1 row in set
  

• Create a cluster-level replicated table dup_t1.

  1. Log in to the sys tenant and create a unit.

     obclient> CREATE RESOURCE UNIT IF NOT EXISTS 2c5g MAX_CPU 2, MEMORY_SIZE '5G';
     Query OK, 0 rows affected
     

  2. Create a resource pool with a unit number of 2.

     obclient> CREATE RESOURCE POOL tenant_pool UNIT = '2c5g', UNIT_NUM = 2, ZONE_LIST = ('z1', 'z2', 'z3');
     Query OK, 0 rows affected
     

  3. Create a user tenant obmysql and specify its locality distribution.

     obclient> CREATE TENANT obmysql resource_pool_list=('tenant_pool'), LOCALITY = "F@z1, F@z2, R@z3", PRIMARY_ZONE = "z1" SET ob_tcp_invited_nodes='%';
     Query OK, 0 rows affected
     

  4. Log in to the obmysql user tenant created in step 3 and create a replicated table.

     obclient> USE test;
     Database changed
     obclient> CREATE TABLE dup_t1(c1 INT) DUPLICATE_SCOPE = 'cluster';
     Query OK, 0 rows affected
     

  5. (Optional) View the broadcast log stream information. The replicated table will be created on this log stream.

     obclient> SELECT * FROM oceanbase.DBA_OB_LS WHERE FLAG LIKE "%DUPLICATE%";
     +-------+--------+--------------+---------------+-------------+---------------------+----------+---------------------+---------------------+-----------+-----------+
     | LS_ID | STATUS | PRIMARY_ZONE | UNIT_GROUP_ID | LS_GROUP_ID | CREATE_SCN          | DROP_SCN | SYNC_SCN            | READABLE_SCN        | FLAG      | UNIT_LIST |
     +-------+--------+--------------+---------------+-------------+---------------------+----------+---------------------+---------------------+-----------+-----------+
     |  1003 | NORMAL | z1;z2        |             0 |           0 | 1683267390195713284 |     NULL | 1683337744205408139 | 1683337744205408139 | DUPLICATE |           |
     +-------+--------+--------------+---------------+-------------+---------------------+----------+---------------------+---------------------+-----------+-----------+
     1 row in set
     

  6. (Optional) View the replica distribution of the replicated table in the sys tenant. The REPLICA_TYPE field indicates the replica type.

     obclient> SELECT * FROM oceanbase.CDB_OB_TABLE_LOCATIONS WHERE TABLE_NAME = "dup_t1";
     +-----------+---------------+------------+----------+------------+----------------+-------------------+------------+---------------+-----------+-------+------+----------------+----------+----------+--------------+-----------------+
     | TENANT_ID | DATABASE_NAME | TABLE_NAME | TABLE_ID | TABLE_TYPE | PARTITION_NAME | SUBPARTITION_NAME | INDEX_NAME | DATA_TABLE_ID | TABLET_ID | LS_ID | ZONE | SVR_IP         | SVR_PORT | ROLE     | REPLICA_TYPE | DUPLICATE_SCOPE |
     +-----------+---------------+------------+----------+------------+----------------+-------------------+------------+---------------+-----------+-------+------+----------------+----------+----------+--------------+-----------------+
     |      1002 | test          | dup_t1     |   500002 | USER TABLE | NULL           | NULL              | NULL       |          NULL |    200001 |  1003 | z1   | 11.xxx.xxx.xxx |    36125 | LEADER   | FULL         | CLUSTER         |
     |      1002 | test          | dup_t1     |   500002 | USER TABLE | NULL           | NULL              | NULL       |          NULL |    200001 |  1003 | z1   | 11.xxx.xxx.xxx  |    36124 | FOLLOWER | READONLY     | CLUSTER         |
     |      1002 | test          | dup_t1     |   500002 | USER TABLE | NULL           | NULL              | NULL       |          NULL |    200001 |  1003 | z2   | 11.xxx.xxx.xxx |    36127 | FOLLOWER | FULL         | CLUSTER         |
     |      1002 | test          | dup_t1     |   500002 | USER TABLE | NULL           | NULL              | NULL       |          NULL |    200001 |  1003 | z2   | 11.xxx.xxx.xxx |    36126 | FOLLOWER | READONLY     | CLUSTER         |
     |      1002 | test          | dup_t1     |   500002 | USER TABLE | NULL           | NULL              | NULL       |          NULL |    200001 |  1003 | z3   | 11.xxx.xxx.xxx |    36128 | FOLLOWER | READONLY     | CLUSTER         |
     |      1002 | test          | dup_t1     |   500002 | USER TABLE | NULL           | NULL              | NULL       |          NULL |    200001 |  1003 | z3   | 11.xxx.xxx.xxx  |    36129 | FOLLOWER | READONLY     | CLUSTER         |
     +-----------+---------------+------------+----------+------------+----------------+-------------------+------------+---------------+-----------+-------+------+----------------+----------+----------+--------------+-----------------+
     6 rows in set
     

  7. Insert and read/write data in the replicated table like a regular table. For a read request, if you use a proxy, the read request may be routed to any OBServer node. If you connect directly to an OBServer node, the read request will be executed on the connected OBServer node as long as the local replica is readable.

     obclient> INSERT INTO dup_t1 VALUES(1);
     Query OK, 1 row affected
     
     obclient> SELECT * FROM dup_t1;
     +------+
     | c1   |
     +------+
     |    1 |
     +------+
     1 row in set
     

• Create a columnstore table tbl1_cg.

  obclient> CREATE TABLE tbl1_cg (col1 INT PRIMARY KEY, col2 VARCHAR(50)) WITH COLUMN GROUP(each column);
  

• Create a table with a columnstore index.

  obclient> CREATE TABLE tbl2_cg (col1 INT PRIMARY KEY, col2 INT, col3 INT, INDEX i1 (col2) WITH COLUMN GROUP(each column));
  

• Create a columnstore table with a columnstore index.

  obclient> CREATE TABLE tbl3_cg (col1 INT PRIMARY KEY, col2 INT, col3 INT, INDEX i1 (col2) WITH COLUMN GROUP(each column)) WITH COLUMN GROUP(each column);
  

• Specify the Skip Index attribute for a column when creating a table.

  obclient> CREATE TABLE test_skidx(
    col1 INT SKIP_INDEX(MIN_MAX, SUM),
    col2 FLOAT SKIP_INDEX(MIN_MAX),
    col3 VARCHAR(1024) SKIP_INDEX(MIN_MAX),
    col4 CHAR(10)
  );
  

• Create a table tbl1 with an integer column col1 and specify that the operation uses 5 parallel threads. Also, specify that the data in the new table tbl1 will be sourced from the query results of table tbl2.

  obclient> CREATE /*+ parallel(5) */ TABLE tbl1 (col1 INT) AS SELECT col1 FROM tbl2;
  

• Create a table tb with a persistent bloom filter at the macroblock level.

  obclient> CREATE TABLE tb(c1 INT PRIMARY KEY, c2 INT) enable_macro_block_bloom_filter = True;
  

• Create a table tbl1 and specify the semi-structured encoding type as encoding and the frequency threshold as 90.

  obclient> CREATE TABLE tbl1 (col1 INT PRIMARY KEY, col2 INT) SEMISTRUCT_PROPERTIES=(ENCODING_TYPE=ENCODING, FREQ_THRESHOLD=90);
  

• Create a table tb and enable the new Flat row storage format (version 2).

  obclient> CREATE TABLE tb(c1 INT PRIMARY KEY, c2 INT) micro_block_format_version = 2;
  

Create a table-level STORAGE_CACHE_POLICY

• Manually specify hot data.

  CREATE TABLE test_table1 (c1 INT, c2 INT) storage_cache_policy (global = "hot");
  

• Specify hot data by time axis.

  CREATE TABLE test_part_func_expr (
  id INT,
  event_time DATETIME NOT NULL,
  data VARCHAR(100))
  STORAGE_CACHE_POLICY (
    BOUNDARY_COLUMN = event_time,
    HOT_RETENTION =1 YEAR
  )
  PARTITION BY RANGE COLUMNS(`event_time`) (
    PARTITION p_prev3h VALUES LESS THAN ("2018-01-01 00:00:00"), 
    PARTITION p_prev2h VALUES LESS THAN ("2019-01-01 00:00:00"), 
    PARTITION p_prev1h VALUES LESS THAN ("2020-01-01 00:00:00"), 
    PARTITION p_current VALUES LESS THAN ("2021-01-01 00:00:00"), 
    PARTITION p_next1h VALUES LESS THAN ("2022-01-01 00:00:00"), 
    PARTITION p_max VALUES LESS THAN MAXVALUE 
  );  
  

Create a partition-level STORAGE_CACHE_POLICY

• Specify the STORAGE_CACHE_POLICY for primary partitions when creating a partitioned table.

  CREATE TABLE tbl3 (col1 INT, col2 INT, col3 INT)
    PARTITION BY RANGE(col1)
    SUBPARTITION BY KEY(col2, col3) SUBPARTITIONS 5
      (PARTITION p0 VALUES LESS THAN(0) STORAGE_CACHE_POLICY = "hot",
       PARTITION p1 VALUES LESS THAN(100) STORAGE_CACHE_POLICY = "auto"
      );
  

• Specify the STORAGE_CACHE_POLICY for subpartitions when creating a partitioned table.

  CREATE TABLE tbl4 (col1 INT, col2 INT)
    PARTITION BY RANGE(col1)
    SUBPARTITION BY RANGE(col2)
      (PARTITION p0 VALUES LESS THAN(100)
          (SUBPARTITION sp0 VALUES LESS THAN(2020) STORAGE_CACHE_POLICY = "hot",
           SUBPARTITION sp1 VALUES LESS THAN(2021)),
      PARTITION p1 VALUES LESS THAN(200)
          (SUBPARTITION sp2 VALUES LESS THAN(2020),
           SUBPARTITION sp3 VALUES LESS THAN(2021) STORAGE_CACHE_POLICY = "hot",
           SUBPARTITION sp4 VALUES LESS THAN(2022))
      );
  

• Specify the STORAGE_CACHE_POLICY for both indexes and partitions when creating a table.

  CREATE TABLE tbl5 (
  c1 INT,
  c2 VARCHAR(50),
  INDEX idx1 (c1) STORAGE_CACHE_POLICY = (GLOBAL = "hot"),
  INDEX idx2 (c2) STORAGE_CACHE_POLICY = (GLOBAL = "auto")
  ) PARTITION BY RANGE (c1)
  (PARTITION p1 VALUES LESS THAN (100) STORAGE_CACHE_POLICY = "hot",
  PARTITION p2 VALUES LESS THAN (200) STORAGE_CACHE_POLICY = "auto",
  PARTITION p3 VALUES LESS THAN MAXVALUE STORAGE_CACHE_POLICY = "none");
  

  Example explanation:

  • When creating the tbl5 table, two indexes are created, each with a different storage cache policy.
  • The table is partitioned by the c1 column using range partitioning, with each partition having a different storage cache policy.
  • The STORAGE_CACHE_POLICY for indexes and partitions can be independently set, allowing for fine-grained storage management.

References

• Create a table
• Change a table