Updated on 2023-01-05 GMT+08:00

From MongoDB to DDS

Supported Source and Destination Databases

Table 1 Supported databases

Source DB

Destination DB

  • On-premises Mongo database (versions 3.2, 3.4, and 4.0)
  • Self-built MongoDB on ECS (versions 3.2, 3.4, and 4.0)
  • MongoDB 3.2, 3.4, and 4.0 on other clouds (Tencent Cloud MongoDB 3.2 is not supported.)
  • DDS DB instances (versions 3.4 and 4.0)
  • DDS DB instance (versions 3.4, 4.0, and 4.2)
    NOTE:

    The destination database version must be the same as or later than the source database version.

Supported Migration Objects

Different types of migration tasks support different migration objects. For details, see Table 2. DRS will automatically check the objects you selected before the migration.

Table 2 Migration objects

Type

Precautions

Migration objects

  • Object level: table level, database level, or instance level (full migration).
  • Supported migration objects:
    • Associated objects must be migrated at the same time to avoid migration failure caused by missing associated objects. Common dependencies: collections referenced by views, and views referenced by views
    • Replica set: Only collections (including validator and capped collections), indexes, and views can be migrated.
    • Cluster: Only collections (including validator and capped collections), shard keys, indexes, and views can be migrated.
    • Single node: Only collections (including validator and capped collections), indexes, and views can be migrated.
    • Only user data and source database account information can be migrated. The system databases (for example, local, admin, and config) and system collection cannot be migrated. If service data is stored in the system database, run the renameCollection command to move the service data to the user database.
    • The statement for creating a view cannot contain a regular expression.
    • Collections that contain the _id field without indexes are not supported.
    • The first parameter of BinData() cannot be 2.
    • If ranged sharding is used, maxKey cannot be used as the primary key.
    NOTE:

    The objects that can be migrated have the following constraints:

    • The source database name cannot contain /\.$ or spaces. The collection name and view name cannot start with system. or contain the dollar sign ($).

Database Account Permission Requirements

To start a migration task, the source and destination database users must have permissions listed in the following table. Different types of migration tasks require different permissions. For details, see Table 3. DRS automatically checks the database account permissions in the pre-check phase and provides handling suggestions.

  • 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 source and destination 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 3 Database account permission

Type

Full Migration

Full+Incremental Migration

Source database user

  • Replica set: The source database user must have the readAnyDatabase permission for the admin database.
  • Single node: The source database user must have the readAnyDatabase permission for the admin database.
  • Cluster: The source database user must have the readAnyDatabase permission for the admin database and the read permission for the config database.
  • To migrate accounts and roles of the source database, the source and destination database users must have the read permission for the system.users and system.roles system tables of the admin database.
  • Replica set: The source database user must have the readAnyDatabase permission for the admin database and the read permission for the local database.
  • Single node: The source database user must have the readAnyDatabase permission for the admin database and the read permission for the local database.
  • Cluster: The source mongos node user must have the readAnyDatabase permission for the admin database and the read permission for the config database. The source shard node user must have the readAnyDatabase permission for the admin database and the read permission for the local database.
  • To migrate accounts and roles of the source database, the source and destination database users must have the read permission for the system.users and system.roles system tables of the admin database.

Destination database user

The user who connects to the destination database must have the dbAdminAnyDatabase permission of the admin database and the readWrite permission of the destination database.

If the destination database is a cluster instance, the database user must have the clusterManager permission for the admin database.

For example, the source database user must have the readAnyDatabase permission for the admin database and the read permission for the config database.

db.grantRolesToUser("Username",[{role:"readAnyDatabase",db:"admin"}, {role:"read",db:"config"}])

Suggestions

  • The success of database migration depends on environment and manual operations. To ensure a smooth migration, perform a migration trial before you start the migration to help you detect and resolve problems in advance.
  • In the migration, ensure that no data is written to the destination database to ensure data consistency before and after the migration.
  • Start your migration task during off-peak hours. A less active database is easier to migrate successfully. If the data is fairly static, there is less likely to be any severe performance impacts during the migration.
    • If network bandwidth is not limited, the query rate of the source database increases by about 20 MB/s during full migration, and two to four CPUs are occupied.
    • 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.
  • 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

The full+incremental migration process consists of four phases: task startup, full synchronization, incremental synchronization, and task completion. A single full migration task contains three phases. To ensure smooth migration, read the following precautions before creating a migration task.

Table 4 Precautions

Type

Restrictions

Starting a task

  • Source database parameter requirements:
    • During an incremental migration, Oplog of the source database must be enabled.
    • If the storage space is sufficient, store the source database Oplog for as long as possible. The recommended retention period is three days.
    • If the incremental source data cluster is migrated, the source database balancer must be disabled.
  • Source database object requirements:
    • The source cannot be a GaussDB(for Mongo) instance.
    • The source database name cannot contain /\.$ or spaces. The collection name and view name cannot start with system. or contain the dollar sign ($).
    • When multiple source databases are migrated to the same destination database, the name of the database to be migrated must be unique.
  • Destination database parameter requirements:
    • The destination DB instance is running properly.
    • The destination DB instance must have sufficient storage space.
    • The version of the destination database instance must be the same as or later than that of the source database.
    • DRS supports full migration between cluster instances. If the source cluster instance is not sharded, ensure that the size of the primary shard on the destination database is greater than that of the source database.
  • Other notes:
    • Associated objects must be migrated at the same time to avoid migration failure caused by missing associated objects. Common dependencies: collections referenced by views, and views referenced by views
    • Replica set: The MongoDB replica set instance must be available and have primary nodes.
    • Source database from a single node instance on other clouds cannot be migrated.
    • If you need to perform an incremental migration for a single node instance, the source database must be a DDS single-node instance on the current cloud.
    • If the source database is not on a cluster instance, the following operations and commands are supported during incremental migration:

      Creating and deleting databases

      Adding, deleting, and updating documents

      Creating and deleting collections

      Creating and deleting indexes

      Creating and deleting views

      The convertToCapped, collMod, and renameCollection commands are supported.

    • During a full plus incremental migration between clusters, the objects to be migrated cannot be deleted. Otherwise, the migration task fails.
    • If you select Cluster (MongoDB 4.0+) for Source DB Instance Type, DRS will use the MongoDB change streams feature during the migration. Note the following before you use change streams:

      Data subscription using change streams consumes a certain amount of CPU and memory resources of the source database. Evaluate the resources of the source database in advance.

      If the load on the source database is heavy, the processing speed of change streams cannot keep up with the oplog generation speed. As a result, DRS synchronization delay occurs.

      Change streams support only the following DDLs: drop database, drop collection and rename

      The DBPointer and DBRef data types are not supported.

      In the incremental migration phase,the migration speed can reach up to 10,000 rows in a single table per second.

      Currently, only whitelisted users can use Change Streams.

    • If you select Oplog for Source Database Type, DRS will create multiple subtasks based on the number of source shards. If Task Rate Limit is selected, the configured rate limit value is synchronized to each subtask.
    • If a Time-to-Live (TTL) index already exists in the collection of the source database or is created during an incremental migration, data consistency cannot be ensured when source and destination databases are in different time zone.
    • The value of block_compressor is determined by stats().wiredTiger.creationString.block_compressor of the collection in the source database. If the destination database contains corresponding empty collections, the compression parameters will not be migrated. If the compression parameters in the source database are not supported by the destination database, configure the compression parameters based on net.compression.compressors of the destination database. If the destination database version is DDS 4.2, DRS does not migrate compression parameters because the destination database does not support compression parameter settings.
    • If the accounts and roles to be migrated conflict with those in the destination database, DRS will skip the conflict data and continue the migration.
    • If the MongoDB service of the source database is deployed with other services on the same server, set the value of the cacheSizeGB parameter to the half of the minimum idle cache for the WiredTiger engine of the source database.
    • If the source is a replica set instance, enter information about all primary and secondary nodes to reduce the impact of a primary/secondary switchover on the migration task. If you enter information about multiple primary and secondary nodes, ensure that all nodes belong to the same replica set instance.
    • If the source is a cluster instance, enter information about multiple mongos nodes to reduce the impact of single-node failure on the migration task. In addition, ensure that all mongos nodes belong to the same cluster instance. For an incremental migration of a cluster instance, you are advised to enter information about all primary and secondary nodes of the shard node and ensure that all node information belongs to the same shard to reduce the impact of a primary/secondary switchover on the migration task. Ensure that all shard nodes belong to the same cluster.
    • During the creation of a migration task, the destination database can be set to read-only or read/write if the DDS kernel version supports this setting. After the task is created, this setting cannot be changed.

      Read-only: During the migration, the destination database is read-only. After the migration is complete, it restores to the read/write status. This option ensures the integrity and success rate of data migration.

      Read/Write: During the migration, the destination database can be queried or modified. Data being migrated may be modified when operations are performed or applications are connected. It should be noted that background processes can often generate or modify data, which may result in data conflicts, task faults, and upload failures. Do not select this option if you do not fully understand the risks.

    • In some migration scenarios, to prevent the drop database operation from deleting the existing collections in the destination database, the drop database operation will not be synchronized to the destination database.

      If the source database version is earlier than MongoDB 3.6, running the drop database command will delete the collections only from the source database. The collections in destination database will not be deleted.

      If the source database version is MongoDB 3.6 or later, the drop database operation is represented by the drop database and drop collection operations in oplog. Running the drop database command will delete the collections from both the source and destination databases.

    • To accelerate the migration, delete unnecessary indexes from the source database and retain only necessary indexes before the migration. You are advised not to create indexes for the source database during the migration. If indexes must be created, create them in the background.
    • To prevent loopback, do not start tasks that migrate the same database to and out of the cloud at the same time.

Full migration

  • During task startup and full migration, do not perform DDL operations on the source database. Otherwise, the task may be abnormal.
  • During migration, do not modify or delete the usernames, passwords, permissions, or ports of the source and destination databases.
  • During migration, do not modify the destination database (including but not limited to DDL and DML operations) that is being migrated.
  • During the migration, data rollback caused by primary/standby switchover of the source database is not supported.
  • During the migration, documents larger than 16 MB in the source database cannot be inserted or updated.

Incremental migration

  • During migration, do not modify or delete the usernames, passwords, permissions, or ports of the source and destination databases.
  • During migration, do not modify the destination database (including but not limited to DDL and DML operations) that is being migrated.
  • During the migration, data rollback caused by primary/standby switchover of the source database is not supported.
  • During the migration, documents larger than 16 MB in the source database cannot be inserted or updated.
  • During an incremental migration of collections, you are advised not to rename the collections.
  • To ensure the migration performance, concurrent replay is performed at the set level in the incremental task phase. In the following scenarios, only single-thread write is supported and concurrent replay is not supported:
    • The collection index contains a unique key.
    • The value of capped of the collection attribute is true.

    In either of the preceding scenarios, the task delay may increase.

Precautions for Comparison

  • You are advised to compare data in the source database during off-peak hours to prevent inconsistent data from being falsely reported and reduce the impact on the source database and DRS tasks.
  • During incremental synchronization, if data is written to the source database, the comparison results may be inconsistent.
  • During row comparison, if an orphan document exists in a cluster instance or chunks are being migrated, the number of returned rows is incorrect and the comparison results are inconsistent. For details, see MongoDB official documents.

Prerequisites

  • You have logged in to the DRS console.
  • For details about the DB types and versions supported by real-time migration, see Real-Time Migration.

Procedure

This section uses MongoDB sharded clusters as an example to describe how to configure a task for migrating MongoDB databases to DDS over a public network.

  1. On the Online Migration Management page, click Create Migration Task.
  2. On the Create Replication Instance page, specify the task name, description, and the replication instance details, and click Create Now.

    • Task information description
      Table 5 Task information

      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 !=<>'&"\

    • Replication instance information
      Table 6 Replication instance settings

      Parameter

      Description

      Data Flow

      Select To the cloud.

      The destination database must be a database on the current cloud.

      Source DB Engine

      Select MongoDB.

      Destination DB Engine

      Select DDS.

      Network Type

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

      • VPC is suitable for migrations between cloud databases of the same account in the same region.
      • Public network is suitable for migrations from on-premises or external cloud databases to the destination databases bound with an EIP.
      • VPN and Direct Connect are suitable for migrations from on-premises databases to cloud databases or between cloud databases across regions.

      Destination DB Instance

      Select the DB instance you have created.

      Replication Instance Subnet

      The subnet where the replication instance resides. 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 there are available IP addresses for the subnet. To ensure that the replication instance is successfully created, only subnets with DHCP enabled are displayed.

      Migration Type

      • Full: This migration type is suitable for scenarios where service interruption is permitted. It migrates all objects and data in non-system databases to the destination database at one time. The objects include collections, views, and indexes.
        NOTE:

        If you are performing a full migration, do not perform operations on the source database. Otherwise, data generated in the source database during the migration will not be synchronized to the destination database.

      • Full+Incremental: This migration type allows you to migrate data without interrupting services. After a full migration initializes the destination database, an incremental migration initiates and parses logs to ensure data consistency between the source and destination databases.
        NOTE:

        If you select Full+Incremental, data generated during the full migration will be continuously synchronized to the destination database, and the source remains accessible.

      Source DB Instance Type

      If you select Full+Incremental for Migration Type, set this parameter based on the source database.

      • If the source database is a cluster instance, set this parameter to Cluster.
      • If the source database is a replica set or a single node instance, set this parameter to Non-cluster.

      Obtain Incremental Data

      This parameter is available for configuration if Source DB Instance Type is set to Cluster. You can determine how to capture data changes during the incremental synchronization.

      • oplog: For MongoDB 3.2 or later, DRS directly connects to each shard of the source DB instance to extract data. If you select this method, you must disable the balancer of the source database. For details, see How Do I Disable the Balancer? When testing the connectivity between the source and the DRS instance, you need to enter the connection information of each shard in the source database on the task configuration page.
      • changeStream: This method is recommended. For MongoDB 4.0 and later, DRS connects to mongos nodes of the source database to extract data. If you select this method, you must enable the WiredTiger storage engine of the source database.
      NOTE:

      Only whitelisted users can use changeStream. To use this function, submit a service ticket. In the upper right corner of the management console, choose Service Tickets > Create Service Ticket to submit a service ticket.

      Destination Database Access

      This option does not appear if the DDS kernel version does not support this setting. This setting cannot be changed after a migration task is created.

      • Read-only

        During migration, the destination database is read-only. After the migration is complete, it restores to the read/write status. This option ensures the integrity and success rate of data migration.

      • Read/Write

        During the migration, the destination database can be queried or modified. Data being migrated may be modified when operations are performed or applications are connected. It should be noted that background processes can often generate or modify data, which may result in data conflicts, task faults, and upload failures. Do not select this option if you do not fully understand the risks. Set the destination database to Read/Write only when you need to modify other data in the database during the migration.

        The task cannot be modified after being created.

      Source Shard Quantity

      If Source DB Instance Type is set to Cluster and Obtain Incremental Data is set to oplog, you need to enter the number of source database shards.

      The number of source shards ranges from 2 to 32. Specify this parameter based on the actual number of shards in the source DB.

    • Enterprise Projects and Tags
      Table 7 Enterprise Projects and Tags

      Parameter

      Description

      Enterprise Project

      An enterprise project you would like to use to centrally manage your Global Accelerator resources. Select an enterprise project from the drop-down list. The default project is default. For more information about enterprise project, see Enterprise Management User Guide.

      Tags

      • Tags a task. This configuration is optional. Adding tags helps you better identify and manage your tasks. Each task can have up to 10 tags.
      • After a task is created, you can view its tag details on the Tags tab. For details, see Tag Management.

    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 Configure Source and Destination Databases page, wait until the replication 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 replication instance. After the connection tests are successful, select the check box before the agreement and click Next.

    • Source database information
      Table 8 Source database information

      Parameter

      Description

      mongos Address

      IP address or domain name of the source database in the IP address/Domain name:Port format. The port of the source database. Range: 1 - 65535

      You can enter a maximum of three groups of IP addresses or domain names of the source database. Separate multiple values with commas (,). For example: 192.168.0.1:8080,192.168.0.2:8080. Ensure that the entered IP addresses or domain names belong to the same sharded cluster.

      NOTE:

      If multiple IP addresses or domain names are entered, the test connection is successful as long as one IP address or domain name is accessible. Therefore, you must ensure that the IP address or domain name is correct.

      Authentication Database

      The name of the authentication database. For example: The default authentication database of DDS instance is admin.

      mongos Username

      The username for accessing the source database.

      mongos Password

      The password for the database username.

      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 not enabled, your data may be at risk.

      Sharded Database

      Enter the information about the sharded databases in the source database.

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

    • Destination database configuration
      Table 9 Destination database settings

      Parameter

      Description

      DB Instance Name

      The DB instance you selected when creating the migration task. This parameter cannot be changed.

      Database Username

      The username for accessing the destination database.

      Database Password

      The password for the database username.

      The username and password of the destination database are encrypted and stored in the database and the replication instance during the migration. After the task is deleted, the username and password are permanently deleted.

  4. On the Set Task page, select migration objects and click Next.

    Table 10 Migrate Object

    Parameter

    Description

    Flow Control

    You can choose whether to control the flow.

    • Yes

      You can customize the maximum migration 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 migration 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 only during a full migration.
      • If you select Cluster for Source DB Instance Type and oplog for Obtain Incremental Data in 2, the flow control settings will be synchronized to each subtask.
      • You can also change the flow control mode after creating a task. For details, see Modifying the Flow Control Mode.

    Migrate Account

    Accounts to be migrated can be classified into the following types: accounts that can be migrated and accounts that cannot be migrated. You can choose whether to migrate the accounts. Accounts that cannot be migrated or accounts that are not selected will not exist in the destination database. Ensure that your services will not be affected by these accounts.
    • No

      During the migration, accounts and roles are not migrated.

    Migrate Object

    The left pane displays the source database objects, and the right pane displays the selected objects. You can choose to migrate all objects, tables, or databases based on your service requirements.

    • All: All objects in the source database are migrated to the destination database. After the migration, the object names will remain the same as those in the source database and cannot be modified.
    • Tables: The selected table-level objects will be migrated.
    • Databases: The selected database-level objects will be migrated.

    If the source database is changed, click in the upper right corner before selecting migration objects to ensure that the objects to be selected are from the changed source database.

    NOTE:
    • If you choose not to migrate all of the databases, the migration may fail because the objects, such as stored procedures and views, in the databases to be migrated may have dependencies on other objects that are not migrated. To prevent migration failure, migrate all of the databases.
    • If the object name contains spaces, the spaces before and after the object name are not displayed. If there are multiple spaces between the object name and the object name, only one space is displayed.
    • The name of the selected migration object cannot contain spaces.
    • To quickly select the desired database objects, you can use the search function.

  5. On the Check Task page, check the migration task.

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

  6. On the displayed page, specify Start Time and confirm that the configured information is correct and click Submit to submit the task.

    Table 11 Task startup settings

    Parameter

    Description

    Started Time

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

    NOTE:

    The migration task may affect the performance of the source and destination databases. You are advised to start the task in off-peak hours and reserve two to three days for data verification.

  7. After the task is submitted, view and manage it on the Online Migration 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 a task in the Configuration state for three days. After three days, DRS automatically deletes background resources, and the task status remains unchanged. When you restart the task configuration, DRS applies for resources again.