使用手册

MySQL 使用手册

环境要求

支持的 MySQL 版本:5.1.15 ~ 8.0.x。

权限要求

  1. 源端开启 binlog

MySQL 5.7 my.cnf 示例配置:

  1. [mysqld]
  2. server-id=1
  3. log-bin=mysql-bin
  4. binlog-format=row
  5. binlog-row-image=full
  6. max_connections=600

执行以下命令,确认是否有开启 binlog:

  1. show variables like '%log_bin%';
  2. show variables like '%binlog%';

如以下显示,则说明 binlog 已开启

  1. +-----------------------------------------+---------------------------------------+
  2. | Variable_name | Value |
  3. +-----------------------------------------+---------------------------------------+
  4. | log_bin | ON |
  5. | binlog_format | ROW |
  6. | binlog_row_image | FULL |
  7. +-----------------------------------------+---------------------------------------+
  1. 赋予源端 MySQL 账号 replication 相关权限。

执行以下命令,查看该用户是否有迁移权限:

  1. SHOW GRANTS FOR 'migration_user';

示例结果:

  1. +------------------------------------------------------------------------------+
  2. |Grants for ${username}@${host} |
  3. +------------------------------------------------------------------------------+
  4. |GRANT REPLICATION SLAVE, REPLICATION CLIENT ON *.* TO ${username}@${host} |
  5. |....... |
  6. +------------------------------------------------------------------------------+
  1. 赋予 MySQL 账号 DDL DML 权限

源端账号需要具备查询权限。 示例:

  1. GRANT SELECT ON migration_ds_0.* TO `migration_user`@`%`;

目标端账号需要具备增删改查等权限。 示例:

  1. GRANT CREATE, DROP, INDEX, SELECT, INSERT, UPDATE, DELETE ON *.* TO `migration_user`@`%`;

详情请参见 MySQL GRANT

完整流程示例

前提条件

  1. 在 MySQL 已准备好源端库、表、数据。
  1. DROP DATABASE IF EXISTS migration_ds_0;
  2. CREATE DATABASE migration_ds_0 DEFAULT CHARSET utf8;
  3. USE migration_ds_0;
  4. CREATE TABLE t_order (order_id INT NOT NULL, user_id INT NOT NULL, status VARCHAR(45) NULL, PRIMARY KEY (order_id));
  5. INSERT INTO t_order (order_id, user_id, status) VALUES (1,2,'ok'),(2,4,'ok'),(3,6,'ok'),(4,1,'ok'),(5,3,'ok'),(6,5,'ok');
  1. 在 MySQL 准备目标端库。
  1. DROP DATABASE IF EXISTS migration_ds_10;
  2. CREATE DATABASE migration_ds_10 DEFAULT CHARSET utf8;
  3. DROP DATABASE IF EXISTS migration_ds_11;
  4. CREATE DATABASE migration_ds_11 DEFAULT CHARSET utf8;
  5. DROP DATABASE IF EXISTS migration_ds_12;
  6. CREATE DATABASE migration_ds_12 DEFAULT CHARSET utf8;

操作步骤

  1. 在 proxy 新建逻辑数据库并配置好存储单元和规则。
  1. CREATE DATABASE sharding_db;
  2. USE sharding_db
  3. REGISTER STORAGE UNIT ds_2 (
  4. URL="jdbc:mysql://127.0.0.1:3306/migration_ds_10?serverTimezone=UTC&useSSL=false",
  5. USER="root",
  6. PASSWORD="root",
  7. PROPERTIES("minPoolSize"="1","maxPoolSize"="20","idleTimeout"="60000")
  8. ), ds_3 (
  9. URL="jdbc:mysql://127.0.0.1:3306/migration_ds_11?serverTimezone=UTC&useSSL=false",
  10. USER="root",
  11. PASSWORD="root",
  12. PROPERTIES("minPoolSize"="1","maxPoolSize"="20","idleTimeout"="60000")
  13. ), ds_4 (
  14. URL="jdbc:mysql://127.0.0.1:3306/migration_ds_12?serverTimezone=UTC&useSSL=false",
  15. USER="root",
  16. PASSWORD="root",
  17. PROPERTIES("minPoolSize"="1","maxPoolSize"="20","idleTimeout"="60000")
  18. );
  19. CREATE SHARDING TABLE RULE t_order(
  20. STORAGE_UNITS(ds_2,ds_3,ds_4),
  21. SHARDING_COLUMN=order_id,
  22. TYPE(NAME="hash_mod",PROPERTIES("sharding-count"="6")),
  23. KEY_GENERATE_STRATEGY(COLUMN=order_id,TYPE(NAME="snowflake"))
  24. );

如果是迁移到异构数据库,那目前需要在 proxy 执行建表语句。

  1. 在 proxy 配置源端存储单元。
  1. REGISTER MIGRATION SOURCE STORAGE UNIT ds_0 (
  2. URL="jdbc:mysql://127.0.0.1:3306/migration_ds_0?serverTimezone=UTC&useSSL=false",
  3. USER="root",
  4. PASSWORD="root",
  5. PROPERTIES("minPoolSize"="1","maxPoolSize"="20","idleTimeout"="60000")
  6. );
  1. 启动数据迁移。
  1. MIGRATE TABLE ds_0.t_order INTO t_order;

或者指定目标端逻辑库:

  1. MIGRATE TABLE ds_0.t_order INTO sharding_db.t_order;
  1. 查看数据迁移作业列表。
  1. SHOW MIGRATION LIST;

示例结果:

  1. +--------------------------------------------+--------------+----------------+--------+---------------------+-----------+
  2. | id | tables | job_item_count | active | create_time | stop_time |
  3. +--------------------------------------------+--------------+----------------+--------+---------------------+-----------+
  4. | j0102p00002333dcb3d9db141cef14bed6fbf1ab54 | ds_0.t_order | 1 | true | 2023-09-20 14:41:32 | NULL |
  5. +--------------------------------------------+--------------+----------------+--------+---------------------+-----------+
  1. 查看数据迁移详情。
  1. SHOW MIGRATION STATUS 'j0102p00002333dcb3d9db141cef14bed6fbf1ab54';

示例结果:

  1. +------+-------------+--------------+--------------------------+--------+-------------------------+-------------------------------+--------------------------+---------------+
  2. | item | data_source | tables | status | active | processed_records_count | inventory_finished_percentage | incremental_idle_seconds | error_message |
  3. +------+-------------+--------------+--------------------------+--------+-------------------------+-------------------------------+--------------------------+---------------+
  4. | 0 | ds_0 | ds_0.t_order | EXECUTE_INCREMENTAL_TASK | true | 6 | 100 | | |
  5. +------+-------------+--------------+--------------------------+--------+-------------------------+-------------------------------+--------------------------+---------------+
  1. 执行数据一致性校验。
  1. CHECK MIGRATION 'j0102p00002333dcb3d9db141cef14bed6fbf1ab54' BY TYPE (NAME='DATA_MATCH');

数据一致性校验算法类型来自:

  1. SHOW MIGRATION CHECK ALGORITHMS;

示例结果:

  1. +-------------+--------------+--------------------------------------------------------------+----------------------------+
  2. | type | type_aliases | supported_database_types | description |
  3. +-------------+--------------+--------------------------------------------------------------+----------------------------+
  4. | CRC32_MATCH | | MySQL,MariaDB,H2 | Match CRC32 of records. |
  5. | DATA_MATCH | | SQL92,MySQL,PostgreSQL,openGauss,Oracle,SQLServer,MariaDB,H2 | Match raw data of records. |
  6. +-------------+--------------+--------------------------------------------------------------+----------------------------+

目标端开启数据加密的情况需要使用DATA_MATCH

异构迁移需要使用DATA_MATCH

查询数据一致性校验进度:

  1. SHOW MIGRATION CHECK STATUS 'j0102p00002333dcb3d9db141cef14bed6fbf1ab54';

示例结果:

  1. +--------------+--------+---------------------+--------+-------------------------------+-----------------------------+--------------------------+-------------------------+-------------------------+------------------+----------------+-----------------+---------------+
  2. | tables | result | check_failed_tables | active | inventory_finished_percentage | inventory_remaining_seconds | incremental_idle_seconds | check_begin_time | check_end_time | duration_seconds | algorithm_type | algorithm_props | error_message |
  3. +--------------+--------+---------------------+--------+-------------------------------+-----------------------------+--------------------------+-------------------------+-------------------------+------------------+----------------+-----------------+---------------+
  4. | ds_0.t_order | true | | false | 100 | 0 | | 2023-09-20 14:45:31.992 | 2023-09-20 14:45:33.519 | 1 | DATA_MATCH | | |
  5. +--------------+--------+---------------------+--------+-------------------------------+-----------------------------+--------------------------+-------------------------+-------------------------+------------------+----------------+-----------------+---------------+
  1. 完成作业。
  1. COMMIT MIGRATION 'j0102p00002333dcb3d9db141cef14bed6fbf1ab54';

更多 DistSQL 请参见 RAL #数据迁移

PostgreSQL 使用手册

环境要求

支持的 PostgreSQL 版本:9.4 或以上版本。

权限要求

  1. 源端开启 test_decoding

  2. 源端调整 WAL 配置。

postgresql.conf 示例配置:

  1. wal_level = logical
  2. max_wal_senders = 10
  3. max_replication_slots = 10
  4. wal_sender_timeout = 0
  5. max_connections = 600

详情请参见 Write Ahead LogReplication

  1. 赋予源端 PostgreSQL 账号 replication 权限。

pg_hba.conf 示例配置:

  1. host replication repl_acct 0.0.0.0/0 md5

详情请参见 The pg_hba.conf File

  1. 赋予源端 PostgreSQL 账号 DDL DML 权限。

如果使用非超级管理员账号进行迁移,要求该账号在迁移时用到的数据库上,具备 CREATE 和 CONNECT 的权限。

示例:

  1. GRANT CREATE, CONNECT ON DATABASE migration_ds_0 TO migration_user;

还需要账号对迁移的表和 schema 具备访问权限,以 test schema 下的 t_order 表为例。

  1. \c migration_ds_0
  2. GRANT USAGE ON SCHEMA test TO GROUP migration_user;
  3. GRANT SELECT ON TABLE test.t_order TO migration_user;

PostgreSQL 有 OWNER 的概念,如果是数据库,SCHEMA,表的 OWNER,则可以省略对应的授权步骤。

详情请参见 PostgreSQL GRANT

完整流程示例

前提条件

  1. 在 PostgreSQL 已准备好源端库、表、数据。
  1. DROP DATABASE IF EXISTS migration_ds_0;
  2. CREATE DATABASE migration_ds_0;
  3. \c migration_ds_0
  4. CREATE TABLE t_order (order_id INT NOT NULL, user_id INT NOT NULL, status VARCHAR(45) NULL, PRIMARY KEY (order_id));
  5. INSERT INTO t_order (order_id, user_id, status) VALUES (1,2,'ok'),(2,4,'ok'),(3,6,'ok'),(4,1,'ok'),(5,3,'ok'),(6,5,'ok');
  1. 在 PostgreSQL 准备目标端库。
  1. DROP DATABASE IF EXISTS migration_ds_10;
  2. CREATE DATABASE migration_ds_10;
  3. DROP DATABASE IF EXISTS migration_ds_11;
  4. CREATE DATABASE migration_ds_11;
  5. DROP DATABASE IF EXISTS migration_ds_12;
  6. CREATE DATABASE migration_ds_12;

操作步骤

  1. 在 proxy 新建逻辑数据库并配置好存储单元和规则。
  1. CREATE DATABASE sharding_db;
  2. \c sharding_db
  3. REGISTER STORAGE UNIT ds_2 (
  4. URL="jdbc:postgresql://127.0.0.1:5432/migration_ds_10",
  5. USER="postgres",
  6. PASSWORD="root",
  7. PROPERTIES("minPoolSize"="1","maxPoolSize"="20","idleTimeout"="60000")
  8. ), ds_3 (
  9. URL="jdbc:postgresql://127.0.0.1:5432/migration_ds_11",
  10. USER="postgres",
  11. PASSWORD="root",
  12. PROPERTIES("minPoolSize"="1","maxPoolSize"="20","idleTimeout"="60000")
  13. ), ds_4 (
  14. URL="jdbc:postgresql://127.0.0.1:5432/migration_ds_12",
  15. USER="postgres",
  16. PASSWORD="root",
  17. PROPERTIES("minPoolSize"="1","maxPoolSize"="20","idleTimeout"="60000")
  18. );
  19. CREATE SHARDING TABLE RULE t_order(
  20. STORAGE_UNITS(ds_2,ds_3,ds_4),
  21. SHARDING_COLUMN=order_id,
  22. TYPE(NAME="hash_mod",PROPERTIES("sharding-count"="6")),
  23. KEY_GENERATE_STRATEGY(COLUMN=order_id,TYPE(NAME="snowflake"))
  24. );

如果是迁移到异构数据库,那目前需要在 proxy 执行建表语句。

  1. 在 proxy 配置源端存储单元。
  1. REGISTER MIGRATION SOURCE STORAGE UNIT ds_0 (
  2. URL="jdbc:postgresql://127.0.0.1:5432/migration_ds_0",
  3. USER="postgres",
  4. PASSWORD="root",
  5. PROPERTIES("minPoolSize"="1","maxPoolSize"="20","idleTimeout"="60000")
  6. );
  1. 启动数据迁移。
  1. MIGRATE TABLE ds_0.t_order INTO t_order;

或者指定目标端逻辑库:

  1. MIGRATE TABLE ds_0.t_order INTO sharding_db.t_order;

也可以指定源端schema:

  1. MIGRATE TABLE ds_0.public.t_order INTO sharding_db.t_order;
  1. 查看数据迁移作业列表。
  1. SHOW MIGRATION LIST;

示例结果:

  1. +--------------------------------------------+--------------+----------------+--------+---------------------+-----------+
  2. | id | tables | job_item_count | active | create_time | stop_time |
  3. +--------------------------------------------+--------------+----------------+--------+---------------------+-----------+
  4. | j0102p00002333dcb3d9db141cef14bed6fbf1ab54 | ds_0.t_order | 1 | true | 2023-09-20 14:41:32 | NULL |
  5. +--------------------------------------------+--------------+----------------+--------+---------------------+-----------+
  1. 查看数据迁移详情。
  1. SHOW MIGRATION STATUS 'j0102p00002333dcb3d9db141cef14bed6fbf1ab54';

示例结果:

  1. +------+-------------+--------------+--------------------------+--------+-------------------------+-------------------------------+--------------------------+---------------+
  2. | item | data_source | tables | status | active | processed_records_count | inventory_finished_percentage | incremental_idle_seconds | error_message |
  3. +------+-------------+--------------+--------------------------+--------+-------------------------+-------------------------------+--------------------------+---------------+
  4. | 0 | ds_0 | ds_0.t_order | EXECUTE_INCREMENTAL_TASK | true | 6 | 100 | | |
  5. +------+-------------+--------------+--------------------------+--------+-------------------------+-------------------------------+--------------------------+---------------+
  1. 执行数据一致性校验。
  1. CHECK MIGRATION 'j0102p00002333dcb3d9db141cef14bed6fbf1ab54';

查询数据一致性校验进度:

  1. SHOW MIGRATION CHECK STATUS 'j0102p00002333dcb3d9db141cef14bed6fbf1ab54';

示例结果:

  1. +--------------+--------+---------------------+--------+-------------------------------+-----------------------------+--------------------------+-------------------------+-------------------------+------------------+----------------+-----------------+---------------+
  2. | tables | result | check_failed_tables | active | inventory_finished_percentage | inventory_remaining_seconds | incremental_idle_seconds | check_begin_time | check_end_time | duration_seconds | algorithm_type | algorithm_props | error_message |
  3. +--------------+--------+---------------------+--------+-------------------------------+-----------------------------+--------------------------+-------------------------+-------------------------+------------------+----------------+-----------------+---------------+
  4. | ds_0.t_order | true | | false | 100 | 0 | | 2023-09-20 14:45:31.992 | 2023-09-20 14:45:33.519 | 1 | DATA_MATCH | | |
  5. +--------------+--------+---------------------+--------+-------------------------------+-----------------------------+--------------------------+-------------------------+-------------------------+------------------+----------------+-----------------+---------------+
  1. 完成作业。
  1. COMMIT MIGRATION 'j0102p00002333dcb3d9db141cef14bed6fbf1ab54';

更多 DistSQL 请参见 RAL #数据迁移

openGauss 使用手册

环境要求

支持的 openGauss 版本:2.0.1 ~ 3.0.0。

权限要求

  1. 调整源端 WAL 配置。

postgresql.conf 示例配置:

  1. wal_level = logical
  2. max_wal_senders = 10
  3. max_replication_slots = 10
  4. wal_sender_timeout = 0
  5. max_connections = 600

详情请参见 Write Ahead LogReplication

  1. 赋予源端 openGauss 账号 replication 权限。

pg_hba.conf 示例配置:

  1. host replication repl_acct 0.0.0.0/0 md5

详情请参见 Configuring Client Access AuthenticationExample: Logic Replication Code

  1. 赋予 openGauss 账号 DDL DML 权限。

如果使用非超级管理员账号进行迁移,要求该账号在迁移时用到的数据库上,具备 CREATE 和 CONNECT 的权限。

示例:

  1. GRANT CREATE, CONNECT ON DATABASE migration_ds_0 TO migration_user;

还需要账号对迁移的表和 schema 具备访问权限,以 test schema 下的 t_order 表为例。

  1. \c migration_ds_0
  2. GRANT USAGE ON SCHEMA test TO GROUP migration_user;
  3. GRANT SELECT ON TABLE test.t_order TO migration_user;

openGauss 有 OWNER 的概念,如果是数据库,SCHEMA,表的 OWNER,则可以省略对应的授权步骤。

openGauss 不允许普通账户在 public schema 下操作。所以如果迁移的表在 public schema 下,需要额外授权。

  1. GRANT ALL PRIVILEGES TO migration_user;

详情请参见 openGauss GRANT

完整流程示例

前提条件

  1. 准备好源端库、表、数据。

1.1. 同构数据库。

  1. DROP DATABASE IF EXISTS migration_ds_0;
  2. CREATE DATABASE migration_ds_0;
  3. \c migration_ds_0
  4. CREATE TABLE t_order (order_id INT NOT NULL, user_id INT NOT NULL, status VARCHAR(45) NULL, PRIMARY KEY (order_id));
  5. INSERT INTO t_order (order_id, user_id, status) VALUES (1,2,'ok'),(2,4,'ok'),(3,6,'ok'),(4,1,'ok'),(5,3,'ok'),(6,5,'ok');

1.2. 异构数据库。

MySQL 示例:

  1. DROP DATABASE IF EXISTS migration_ds_0;
  2. CREATE DATABASE migration_ds_0 DEFAULT CHARSET utf8;
  3. USE migration_ds_0;
  4. CREATE TABLE t_order (order_id INT NOT NULL, user_id INT NOT NULL, status VARCHAR(45) NULL, PRIMARY KEY (order_id));
  5. INSERT INTO t_order (order_id, user_id, status) VALUES (1,2,'ok'),(2,4,'ok'),(3,6,'ok'),(4,1,'ok'),(5,3,'ok'),(6,5,'ok');
  1. 在 openGauss 准备目标端库。
  1. DROP DATABASE IF EXISTS migration_ds_10;
  2. CREATE DATABASE migration_ds_10;
  3. DROP DATABASE IF EXISTS migration_ds_11;
  4. CREATE DATABASE migration_ds_11;
  5. DROP DATABASE IF EXISTS migration_ds_12;
  6. CREATE DATABASE migration_ds_12;

操作步骤

  1. 在 proxy 新建逻辑数据库并配置好存储单元和规则。

1.1. 创建逻辑库。

  1. CREATE DATABASE sharding_db;
  2. \c sharding_db

1.2. 注册存储单元。

  1. REGISTER STORAGE UNIT ds_2 (
  2. URL="jdbc:opengauss://127.0.0.1:5432/migration_ds_10",
  3. USER="gaussdb",
  4. PASSWORD="Root@123",
  5. PROPERTIES("minPoolSize"="1","maxPoolSize"="20","idleTimeout"="60000")
  6. ), ds_3 (
  7. URL="jdbc:opengauss://127.0.0.1:5432/migration_ds_11",
  8. USER="gaussdb",
  9. PASSWORD="Root@123",
  10. PROPERTIES("minPoolSize"="1","maxPoolSize"="20","idleTimeout"="60000")
  11. ), ds_4 (
  12. URL="jdbc:opengauss://127.0.0.1:5432/migration_ds_12",
  13. USER="gaussdb",
  14. PASSWORD="Root@123",
  15. PROPERTIES("minPoolSize"="1","maxPoolSize"="20","idleTimeout"="60000")
  16. );

1.3. 创建分片规则。

  1. CREATE SHARDING TABLE RULE t_order(
  2. STORAGE_UNITS(ds_2,ds_3,ds_4),
  3. SHARDING_COLUMN=order_id,
  4. TYPE(NAME="hash_mod",PROPERTIES("sharding-count"="6")),
  5. KEY_GENERATE_STRATEGY(COLUMN=order_id,TYPE(NAME="snowflake"))
  6. );

1.4. 创建目标端表。

如果是迁移到异构数据库,那目前需要在 proxy 执行建表语句。

  1. CREATE TABLE t_order (order_id INT NOT NULL, user_id INT NOT NULL, status VARCHAR(45) NULL, PRIMARY KEY (order_id));
  1. 在 proxy 配置源端存储单元。

2.1. 同构数据库。

  1. REGISTER MIGRATION SOURCE STORAGE UNIT ds_0 (
  2. URL="jdbc:opengauss://127.0.0.1:5432/migration_ds_0",
  3. USER="gaussdb",
  4. PASSWORD="Root@123",
  5. PROPERTIES("minPoolSize"="1","maxPoolSize"="20","idleTimeout"="60000")
  6. );

2.2. 异构数据库。

MySQL 示例:

  1. REGISTER MIGRATION SOURCE STORAGE UNIT ds_0 (
  2. URL="jdbc:mysql://127.0.0.1:3306/migration_ds_0?serverTimezone=UTC&useSSL=false",
  3. USER="root",
  4. PASSWORD="root",
  5. PROPERTIES("minPoolSize"="1","maxPoolSize"="20","idleTimeout"="60000")
  6. );
  1. 启动数据迁移。
  1. MIGRATE TABLE ds_0.t_order INTO t_order;

或者指定目标端逻辑库:

  1. MIGRATE TABLE ds_0.t_order INTO sharding_db.t_order;

也可以指定源端schema:

  1. MIGRATE TABLE ds_0.public.t_order INTO sharding_db.t_order;
  1. 查看数据迁移作业列表。
  1. SHOW MIGRATION LIST;

示例结果:

  1. +--------------------------------------------+--------------+----------------+--------+---------------------+-----------+
  2. | id | tables | job_item_count | active | create_time | stop_time |
  3. +--------------------------------------------+--------------+----------------+--------+---------------------+-----------+
  4. | j0102p00002333dcb3d9db141cef14bed6fbf1ab54 | ds_0.t_order | 1 | true | 2023-09-20 14:41:32 | NULL |
  5. +--------------------------------------------+--------------+----------------+--------+---------------------+-----------+
  1. 查看数据迁移详情。
  1. SHOW MIGRATION STATUS 'j0102p00002333dcb3d9db141cef14bed6fbf1ab54';

示例结果:

  1. +------+-------------+--------------+--------------------------+--------+-------------------------+-------------------------------+--------------------------+---------------+
  2. | item | data_source | tables | status | active | processed_records_count | inventory_finished_percentage | incremental_idle_seconds | error_message |
  3. +------+-------------+--------------+--------------------------+--------+-------------------------+-------------------------------+--------------------------+---------------+
  4. | 0 | ds_0 | ds_0.t_order | EXECUTE_INCREMENTAL_TASK | true | 6 | 100 | | |
  5. +------+-------------+--------------+--------------------------+--------+-------------------------+-------------------------------+--------------------------+---------------+
  1. 执行数据一致性校验。
  1. CHECK MIGRATION 'j0102p00002333dcb3d9db141cef14bed6fbf1ab54';

查询数据一致性校验进度:

  1. SHOW MIGRATION CHECK STATUS 'j0102p00002333dcb3d9db141cef14bed6fbf1ab54';

示例结果:

  1. +--------------+--------+---------------------+--------+-------------------------------+-----------------------------+--------------------------+-------------------------+-------------------------+------------------+----------------+-----------------+---------------+
  2. | tables | result | check_failed_tables | active | inventory_finished_percentage | inventory_remaining_seconds | incremental_idle_seconds | check_begin_time | check_end_time | duration_seconds | algorithm_type | algorithm_props | error_message |
  3. +--------------+--------+---------------------+--------+-------------------------------+-----------------------------+--------------------------+-------------------------+-------------------------+------------------+----------------+-----------------+---------------+
  4. | ds_0.t_order | true | | false | 100 | 0 | | 2023-09-20 14:45:31.992 | 2023-09-20 14:45:33.519 | 1 | DATA_MATCH | | |
  5. +--------------+--------+---------------------+--------+-------------------------------+-----------------------------+--------------------------+-------------------------+-------------------------+------------------+----------------+-----------------+---------------+
  1. 完成作业。
  1. COMMIT MIGRATION 'j0102p00002333dcb3d9db141cef14bed6fbf1ab54';

更多 DistSQL 请参见 RAL #数据迁移