Help Center/ Data Replication Service/ User Guide (Ankara Region)/ Real-Time Disaster Recovery/ DR Scenarios/ From MySQL to MySQL (Dual-Active DR)
Updated on 2024-04-11 GMT+08:00
View PDF
Share

From MySQL to MySQL (Dual-Active DR)

Supported Source and Destination Databases

Table 1 Supported databases

Service database

DR Database
  • On-premises MySQL databases
  • MySQL databases on an ECS
  • MySQL databases on other clouds
  • RDS for MySQL
  • RDS for MySQL

Prerequisites

Suggestions

  • During the DR initialization, do not perform DDL operations on the service database. Otherwise, the task may be abnormal.
  • During DR initialization, ensure that no data is written to the DR database to ensure data consistency before and after DR.
  • The success of DR depends on environment and manual operations. To ensure a smooth DR, perform a DR trial before you start the DR task to help you detect and resolve problems in advance.
  • It is recommended that you start your DR task during off-peak hours to minimize the impact on your services.
    • If the bandwidth is not limited, initialization of DR will increase query workload of the source database by 50 MB/s and occupy 2 to 4 vCPUs.
    • To ensure data consistency, tables without a primary key may be locked for 3s during disaster recovery.
    • The data in the DR process may be locked by other transactions for a long period of time, resulting in read timeout.
    • If DRS concurrently reads data from a database, it will use about 6 to 10 sessions. The impact of the connections on services must be considered.
    • If you read a table, especially a large table, during DR, the exclusive lock on that table may be blocked.
  • Data-Level Comparison

    To obtain accurate comparison results, start data comparison at a specified time point during off-peak hours. If it is needed, select Start at a specified time for Comparison Time. Due to slight time difference and continuous operations on data, data inconsistency may occur, reducing the reliability and validity of the comparison results.

Precautions

Before creating a DR task, read the following precautions:

  • You are advised to create an independent database account for DRS task connection to prevent task failures caused by database account password modification.
  • After changing the account passwords for the service or DR databases, modify the connection information in the DRS task as soon as possible to prevent automatic retry after a task failure. Automatic retry will lock the database accounts.
Table 2 Precautions

Type

Restrictions

Database permissions
  • The service database user must have the following permissions: SELECT, CREATE, ALTER, DROP, DELETE, INSERT, UPDATE, TRIGGER, REFERENCES, SHOW VIEW, EVENT, INDEX, LOCK TABLES, CREATE VIEW, CREATE ROUTINE, ALTER ROUTINE, CREATE USER, RELOAD, REPLICATION SLAVE, REPLICATION CLIENT, and WITH GRANT OPTION.
  • The DR database user must have the following permissions: SELECT, CREATE, ALTER, DROP, DELETE, INSERT, UPDATE, TRIGGER, REFERENCES, SHOW VIEW, EVENT, INDEX, LOCK TABLES, CREATE VIEW, CREATE ROUTINE, ALTER ROUTINE, CREATE USER, RELOAD, REPLICATION SLAVE, REPLICATION CLIENT, and WITH GRANT OPTION.
  • The root account of the RDS for MySQL DB instance has the preceding permissions by default.

Disaster recovery objects
  • Tables with storage engine different to MyISAM and InnoDB do not support disaster recovery.
  • System tables are not supported.
  • Triggers and events do not support disaster recovery.
  • Accounts that have operation permissions on customized objects in the system database cannot be used for disaster recovery.
  • DDL operations cannot be executed on the active database 2.

Service database configuration
  • The binlog of the MySQL service database must be enabled and use the row-based format.
  • If the storage space is sufficient, store the service database binlog for as long as possible. The recommended retention period is seven days.
    • For self-built MySQL databases, you can set the expire_logs_days parameter to specify the binlog retention period.
    • If the source database is an RDS for MySQL instance, set the binlog retention period by following the instructions provided in RDS User Guide.
  • The service database username or password cannot be empty.
  • server_id in the MySQL service database must be set. If the service database version is MySQL 5.6 or earlier, the server_id value ranges from 2 to 4294967296. If the service database is MySQL 5.7 or later, the server_id value ranges from 1 to 4294967296.
  • During disaster recovery, if the session variable character_set_client is set to binary, some data may include garbled characters.
  • GTID must be enabled for the database.
  • The service database name must contain 1 to 64 characters, including only lowercase letters, digits, hyphens (-), and underscores (_).
  • The table name and view name in the service database cannot contain non-ASCII characters, or the following characters: '<>/\
  • If the expire_logs_days value of the service database is set to 0, the disaster recovery may fail.
  • If tables that have no primary key contain hidden primary keys in the service database, the DR task may fail or data may be inconsistent.

DR database configuration
  • The DR DB instance is running properly. If the DR DB instance is a primary/standby instance, the replication status must also be normal.
  • The DR DB instance must have sufficient storage space.
  • The major version of the active database 1 must be the same as that of the active database 2.
  • The binlog of the DR database must be enabled and use the row-based format.
  • GTID must be enabled for the DR database.
  • In addition to the MySQL system database, the active database 2 must be an empty instance. After the forward task is started, active database 2 is set to read-only. After the backward task is started and DR is performed, the active database 2 is restored to read-write.

Precautions
  • Dual-active DR supports backup in backward and forward directions. Due to certain uncontrollable factors, data may be inconsistent between the two sides. For example, if the load of active database 1 is too heavy and the load of active database 2 is light, data updates on the active database 1 synchronized to the active database 2 will be delayed due to the heave load, as a result, the operation sequence is changed and data becomes inconsistency. Therefore, divide data by unit (database, table, or row) and ensure the unit on one database is responsible for data read and write while on the other is read-only. In essence, in dual-active DR, both the databases play the active role but work differently. For details about common scenarios, see Common Exceptions in Real-Time Disaster Recovery.
  • Before creating a DRS task, if concurrency control rules of SQL statements are configured for the service or DR database, the DRS task may fail.
  • During the DR initialization, do not perform DDL operations on the source database. Otherwise, the DR task may be abnormal.
  • During disaster recovery, you can create accounts for the service database.
  • If the same data on both databases is updated simultaneously, data conflicts may occur. DRS resolves the conflict by overwriting the previous settings with the last settings.
    • When the deletion operation is performed, data is deleted and DRS does not perform any operation.
    • When the insert operation is performed, DRS updates data with the latest inserted data.
    • When the update operation is performed, the original data has been updated and DRS directly insert the new data.
  • Primary key conflicts between the two sides need to be avoided. For example, you can use a UUID or the primary key rule of region+auto-increment ID to avoid conflicts.
  • If the synchronization delay takes a long time due to connection interruption or network issues, you need to determine whether your services can tolerant the long-term delay.
  • Cascade operations cannot be performed on tables with foreign keys. If the foreign key index of a table is a common index, the table structure may fail to be created. You are advised to use a unique index.
  • If a physically generated column in a table is generated based on a time type, the data in the column may be inconsistent.
  • The dual-active DR is different from the single-active DR. Therefore, no active/standby switchover is required.
  • The DR latency is uncontrollable. Therefore, DDL operations must be performed when no service is running, and both RPO and RTO are zero and latency is kept within 30 seconds on active database 1. Do not perform DDL operations on active database 2. (DRS synchronizes only the DDL operations on active database 1 to active database 2.)
  • Ensure that the tables, columns, and rows are consistent in both the databases. (The table structures of both the active databases are consistent.)
  • A backward task can be started only when the forward task is in the DR process and both RPO and RTO are less than 60s.
  • After the dual-active DR task is in the DR process, perform tests on the active database 2 first. If the test results meet the requirements, switch certain service traffic to the active database 2.

Procedure

  1. On the Disaster Recovery Management page, click Create Disaster Recovery Task.
  2. On the Create Disaster Recovery Instance page, specify the task name, description, and the DR instance details, and click Create Now.

    • Task information description
      Table 3 Task and recipient description

      Parameter

      Description

      Task Name

      The task name must start with a letter and consist of 4 to 50 characters. It can contain only letters, digits, hyphens (-), and underscores (_).

      Description

      The description consists of a maximum of 256 characters and cannot contain special characters !=<>'&"\
    • DR instance information
      Table 4 DR instance settings

      Parameter

      Description

      DR Type

      Select Dual-active.

      The DR type can be single-active or dual-active. If Dual-active is selected, two subtasks are created by default, a forward DR task and a backward DR task.

      Current Cloud RDS Instance Role

      Select Active 1 or Active 2. This parameter specifies the role of the current RDS DB instance in the DR relationship and is available when DR Type is set to Dual-active. For details, see How Do I Select Active Database 1 and 2 for Dual-Active DR?

      • Active 1: Initial data is available on the current cloud RDS when a task is created.
      • Active 2: The RDS DB instance on the current cloud is empty when a task is created.

      Active 2 is used as an example.

      Service DB Engine

      Select MySQL.

      DR DB Engine

      Select MySQL.

      Network Type

      The public network is used as an example.

      Available options: VPN or Direct Connect and Public network. By default, the value is Public network.

      DR DB Instance

      The RDS for MySQL instance you created.

      Disaster Recovery Instance Subnet

      Select the subnet where the disaster recovery instance is located. You can also click View Subnet to go to the network console to view the subnet where the instance resides.

      By default, the DRS instance and the destination DB instance are in the same subnet. You need to select the subnet where the DRS instance resides and ensure that there are available IP addresses. To ensure that the DR instance can be successfully created, only subnets with DHCP enabled are displayed.

    If a task fails to be created, DRS retains the task for three days by default. After three days, the task automatically ends.

  3. On the Disaster Recovery Management page, after the task is created, locate the forward subtask and click Edit in the Operation column. The Configure Source and Destination Databases page is displayed.
  4. On the Configure Source and Destination Databases page, wait until the DR instance is created. Then, specify source and destination database information and click Test Connection for both the source and destination databases to check whether they have been connected to the DR instance. After the connection tests are successful, select the check box before the agreement and click Next.

    Table 5 Service database settings

    Parameter

    Description

    Source Database Type

    By default, Self-built on ECS is selected.

    IP Address or Domain Name

    The IP address or domain name of the service database.

    Port

    The port of the service database. Range: 1 – 65535

    Database Username

    The username for accessing the service database.

    Database Password

    The password for the service database username. You can change the password if necessary. To change the password, perform the following operation after the task is created:

    If the task is in the Starting, Initializing, Disaster recovery in progress, or Disaster recovery failed status, in the Connection Information area on the Basic Information tab, click Modify Connection Details. In the displayed dialog box, change the password.

    SSL Connection

    SSL encrypts the connections between the source and destination databases. If SSL is enabled, upload the SSL CA root certificate.

    NOTE:
    • The maximum size of a single certificate file that can be uploaded is 500 KB.
    • If SSL is disabled, your data may be at risk.

    The IP address, domain name, username, and password of the service database are encrypted and stored in DRS and will be cleared after the task is deleted.

    Table 6 DR database settings

    Parameter

    Description

    DB Instance Name

    The RDS for MySQL instance you selected when you create the DR instance. The instance name cannot be changed.

    Database Username

    The username for accessing the DR database.

    Database Password

    The password for the database username. The password can be changed after a task is created.

    If the task is in the Starting, Initializing, Disaster recovery in progress, or Disaster recovery failed status, in the Connection Information area on the Basic Information tab, click Modify Connection Details. In the displayed dialog box, change the password.

    The database username and password are encrypted and stored in DRS, and will be cleared after the task is deleted.

    SSL Connection

    If SSL connection is required, enable SSL on the DR database, ensure that related parameters have been correctly configured, and upload an SSL certificate.

    NOTE:
    • The maximum size of a single certificate file that can be uploaded is 500 KB.
    • If SSL is disabled, your data may be at risk.

  5. On the Configure DR page, specify flow control and click Next.

    Table 7 DR settings

    Parameter

    Description

    Flow Control

    You can choose whether to control the flow.

    • Yes

      You can customize the maximum DR speed.

      In addition, you can set the time range based on your service requirements. The traffic rate setting usually includes setting of a rate limiting time period and a traffic rate value. Flow can be controlled all day or during specific time ranges. The default value is All day. A maximum of three time ranges can be set, and they cannot overlap.

      The flow rate must be set based on the service scenario and cannot exceed 9,999 MB/s.

    • No
      The DR speed is not limited and the outbound bandwidth of the source database is maximally used, which causes read consumption on the source database accordingly. For example, if the outbound bandwidth of the source database is 100 MB/s and 80% bandwidth is used, the I/O consumption on the source database is 80 MB/s.
      NOTE:
      • Flow control mode takes effect during the initial DR phase only.
      • You can also change the flow control mode when the task is in the Configuration state. On the Basic Information tab, In the DR Information area, click Modify next to Flow Control. In the dialog box that is displayed, change the flow control mode. The flow control mode cannot be changed for a task that is in Starting state.

    Migrate Definer to User

    Indicates whether to migrate the Definers of all source database objects to the destination database user entered during the connection test.

  6. On the Check Task page, check the DR task.

    • If any check fails, review the failure cause and rectify the fault. After the fault is rectified, click Check Again.

      For details about how to handle check items that fail to pass the pre-check, see Solutions to Failed Check Items.

    • If the check is complete and the check success rate is 100%, click Next.

      You can proceed to the next step only when all checks are successful. If there are any items that require confirmation, view and confirm the details first before proceeding to the next step.

  7. On the displayed page, specify Start Time for the forward subtask. After confirming that the configured information is correct, click Submit to submit the forward DR task.

    Table 8 Task and recipient description

    Parameter

    Description

    Start Time

    Set Start Time to Start upon task creation or Start at a specified time based on site requirements.

    NOTE:

    After a DR task is started, the performance of the service and DR databases may be affected. You are advised to start a DR task during off-peak hours.

  8. Return to the Disaster Recovery Management page. After the forward subtask enters the Disaster recovery in progress state, locate the backward subtask and click Edit in the Operation column. The Configure Source and Destination Databases page of the backward subtask is displayed.

  9. On the Configure Source and Destination Databases page, click Test Connection for both the source and destination databases to check whether they have been connected to the DR instance. After the connection tests are successful, click Next.
  10. On the displayed page, specify Start Time for the backward subtask. After confirming that the configured information is correct, click Submit to submit the backward DR task.
  11. After the task is submitted, view and manage it on the Disaster Recovery Management page.

    • You can view the task status. For more information about task status, see Task Statuses.
    • You can click in the upper-right corner to view the latest task status.
    • By default, DRS retains any task in the Configuration state for three days. After three days, DRS automatically deletes background resources, but the task status remains unchanged. When you reconfigure the task, DRS applies for resources again.

Parent topic: DR Scenarios