Migrating Data from HBase to MRS

Scenario

This section describes how to migrate data from offline IDCs or public cloud HBase clusters to Huawei Cloud MRS. The data volume can be tens of TBs or less. This section uses Huawei Cloud CDM 2.9.1.200 as an example to describe how to migrate data.

Figure 1 HBase data migration

HBase stores data in HDFS, including HFile and WAL files. The hbase.rootdir configuration item specifies the HDFS path. By default, data is stored in the /hbase folder on MRS.

Some mechanisms and tool commands of HBase can also be used to migrate data. For example, you can migrate data by exporting snapshots, exporting/importing data, and CopyTable. For details, see the Apache official website.

This document describes how to migrate HBase data using Huawei Cloud CDM.

For details about the data sources supported by CDM, see Supported Data Sources. If the data source is Apache HBase, the recommended version is 2.1.X or 1.3.X. Before performing the migration, ensure that the data source supports migration.

Solution Advantages

Scenario-based migration migrates snapshots and then restores table data to speed up migration.

Impact on the System

Migrating large volumes of data has high requirements on network communication. When a migration task is executed, other services may be affected. You are advised to migrate data during off-peak hours.

Full Data Migration

1. Log in to the CDM management console.
2. Create a CDM cluster. The security group, VPC, and subnet of the CDM cluster must be the same as those of the destination cluster to ensure that the CDM cluster can communicate with the MRS cluster.
3. On the Cluster Management page, locate the row containing the desired cluster and click Job Management in the Operation column.
4. On the Links tab page, click Create Link.
5. Create a link to the source cluster by referring to Creating a CDM Link. Select a connector type based on the actual cluster, for example, Apache HBase.
   

   (Optional) Use a user with high permissions to migrate HBase. For example, click Show Advanced Attributes and add user hadoop.user.name = Username (for example, omm).

   Figure 2 Link to the source cluster
   

6. On the Links tab page, click Create Link.
7. Create a link to the destination cluster by referring to Creating a CDM Link. Select a connector type based on the actual cluster, for example, MRS HBase.
   

   (Optional) Use a user with high permissions to migrate HBase. For example, click Show Advanced Attributes and add user hadoop.user.name = Username (for example, omm).

   Figure 3 Link to the destination cluster
   

8. Choose Job Management and click the Table/File Migration tab. Then, click Create Job.
9. In the job creation dialog box, configure the job name, source job parameters, and destination job parameters, select the data table to be migrated, and click Next.
   Figure 4 HBase job configuration
   

10. Configure the mapping between the source fields and destination fields and click Next.
11. On the task configuration page that is displayed, click Save without any modification.
12. Choose Job Management and click Table/File Migration. Locate the row containing the job to run and click Run in the Operation column to start migrating HBase data.
13. After the migration is complete, you can run the same query statement in the source and destination clusters to compare the query results.

    Example:

    • Query the number of records in the BTable table on the source and destination clusters to check whether the number of data records is the same. Add the --endtime parameter to eliminate the impact of data updates on the source cluster during the migration.

      Hbase org.apache.hadoop.hbase.mapreduce.RowCounter BTable --endtime=1587973835000

      Figure 5 Querying the number of records in the BTable table
      
    • Use scan ' BTable ', {TIMERANGE=>[1587973235000, 1587973835000]} of HBase shell to query data in a specified period for comparison.

Incremental Data Migration

If new data exists in the source cluster before the service cutover, you need to periodically migrate the new data to the destination cluster. Generally, the data volume updated every day is at the GB level. You can use the Entire DB migration function of CDM to migrate new HBase data every day.

If the Entire DB Migration function of CDM is used, the deleted data in the source HBase cluster cannot be synchronized to the destination cluster.

The HBase connector for scenario migration cannot be shared with that for entire database migration. Therefore, a new HBase connector is required.

1. Repeat 1 to 7 in Full Data Migration to create two HBase connectors. Select MRS HBase and Apache HBase as the connector type for the source cluster and destination cluster, respectively.
   Figure 6 HBase incremental migration link
   

2. Choose Job Management > Entire DB Migration, and click Create Job.
3. On the job parameter configuration page, configure job parameters and click Next.
   • Job Name: Enter a user-defined job name, for example, hbase-increase.
   • Source Job Configuration: Set Source Link Name to the name of the link to the source cluster created in 1, and click Show Advanced Attributes to configure the time range for data migration.
   • Destination Job Configuration: Set Destination Link Name to the name of the link to the destination cluster created in 1. Leave other parameters blank.
   Figure 7 HBase incremental migration job configuration
   

4. Select the data table to be migrated, click Next, and click Save.
5. Choose Job Management > Entire DB Migration and click Run in the Operation column of the job to be executed to start HBase incremental data migration.