BR Tool Overview

BR (Backup & Restore) is a command-line tool for distributed backup and restoration of the TiDB cluster data.

Compared with Dumpling, BR is more suitable for scenarios involved huge data volumes.

In addition to regular backup and restoration, you can also use BR for large-scale data migration as long as compatibility is ensured.

This document describes BR's implementation principles, recommended deployment configuration, usage restrictions and several methods to use BR.

Implementation principles

BR sends the backup or restoration commands to each TiKV node. After receiving these commands, TiKV performs the corresponding backup or restoration operations.

Each TiKV node has a path in which the backup files generated in the backup operation are stored and from which the stored backup files are read during the restoration.

BackupRequest{
    ClusterId,      // The cluster ID.
    StartKey,       // The starting key of the backup (backed up).
    EndKey,         // The ending key of the backup (not backed up).
    StartVersion,   // The version of the last backup snapshot, used for the incremental backup.
    EndVersion,     // The backup snapshot time.
    StorageBackend, // The path where backup files are stored.
    RateLimit,      // Backup speed (MB/s).
}

Deploy and use BR

Recommended deployment configuration

• It is recommended that you deploy BR on the PD node.
• It is recommended that you mount a high-performance SSD to BR nodes and all TiKV nodes. A 10-gigabit network card is recommended. Otherwise, bandwidth is likely to be the performance bottleneck during the backup and restore process.

Usage restrictions

The following are the limitations of using BR for backup and restoration:

• When BR restores data to the upstream cluster of TiCDC/Drainer, TiCDC/Drainer cannot replicate the restored data to the downstream.
• BR supports operations only between clusters with the same new_collations_enabled_on_first_bootstrap value because BR only backs up KV data. If the cluster to be backed up and the cluster to be restored use different collations, the data validation fails. Therefore, before restoring a cluster, make sure that the switch value from the query result of the select VARIABLE_VALUE from mysql.tidb where VARIABLE_NAME='new_collation_enabled'; statement is consistent with that during the backup process.

Compatibility

The compatibility issues of BR and the TiDB cluster are divided into the following categories:

• Some versions of BR are not compatible with the interface of the TiDB cluster.
• The KV format might change when some features are enabled or disabled. If these features are not consistently enabled or disabled during backup and restore, compatibility issues might occur.

These features are as follows:

However, even after you have ensured that the above features are consistently enabled or disabled during backup and restore, compatibility issues might still occur due to the inconsistent internal versions or inconsistent interfaces between BR and TiKV/TiDB/PD. To avoid such cases, BR has the built-in version check.

Version check

Before performing backup and restore, BR compares and checks the TiDB cluster version and the BR version. If there is a major-version mismatch (for example, BR v4.x and TiDB v5.x), BR prompts a reminder to exit. To forcibly skip the version check, you can set --check-requirements=false.

Note that skipping the version check might introduce incompatibility. The version compatibility information between BR and TiDB versions are as follows:

Back up and restore table data in the mysql system schema (experimental feature)

Before v5.1.0, BR filtered out data from the system schema mysql during the backup. Since v5.1.0, BR backs up all data by default, including the system schemas mysql.*. But the technical implementation of restoring the system tables in mysql.* is not complete yet, so the tables in the system schema mysql are not restored by default.

If you want the data of a system table (for example, mysql.usertable1) to be restored to the system schema mysql, you can set the filter parameter to filter the table name (-f "mysql.usertable1"). After the setting, the system table is first restored to the temporary schema, and then to the system schema through renaming.

It should be noted that the following system tables cannot be restored correctly due to technical reasons. Even if -f "mysql.*" is specified, these tables will not be restored:

• Tables related to statistics: "stats_buckets", "stats_extended", "stats_feedback", "stats_fm_sketch", "stats_histograms", "stats_meta", "stats_top_n"
• Tables related to privileges or the system: "tidb", "global_variables", "columns_priv", "db", "default_roles", "global_grants", "global_priv", "role_edges", "tables_priv", "user", "gc_delete_range", "Gc_delete_range_done", "schema_index_usage"

Minimum machine configuration required for running BR

The minimum machine configuration required for running BR is as follows:

In general scenarios (less than 1000 tables for backup and restore), the CPU consumption of BR at runtime does not exceed 200%, and the memory consumption does not exceed 4 GB. However, when backing up and restoring a large number of tables, BR might consume more than 4 GB of memory. In a test of backing up 24000 tables, BR consumes about 2.7 GB of memory, and the CPU consumption remains below 100%.

Best practices

The following are some recommended operations for using BR for backup and restoration:

• It is recommended that you perform the backup operation during off-peak hours to minimize the impact on applications.
• BR supports restore on clusters of different topologies. However, the online applications will be greatly impacted during the restore operation. It is recommended that you perform restore during the off-peak hours or use rate-limit to limit the rate.
• It is recommended that you execute multiple backup operations serially. Running different backup operations in parallel reduces backup performance and also affects the online application.
• It is recommended that you execute multiple restore operations serially. Running different restore operations in parallel increases Region conflicts and also reduces restore performance.
• It is recommended that you mount a shared storage (for example, NFS) on the backup path specified by -s, to make it easier to collect and manage backup files.
• It is recommended that you use a storage hardware with high throughput, because the throughput of a storage hardware limits the backup and restoration speed.
• It is recommended that you disable the checksum feature (--checksum = false) during backup operation and only enable it during the restore operation to reduce migration time. This is because BR by default respectively performs checksum calculation after backup and restore operations to compare the stored data with the corresponding cluster data to ensure accuracy.

How to use BR

Currently, the following methods are supported to run the BR tool:

• Use SQL statements
• Use the command-line tool
• Use BR In the Kubernetes environment

Use SQL statements

TiDB supports both BACKUP and RESTORE SQL statements. The progress of these operations can be monitored with the statement SHOW BACKUPS|RESTORES.

Use the command-line tool

The br command-line utility is available as a separate download. For details, see Use BR Command-line for Backup and Restoration.

In the Kubernetes environment

In the Kubernetes environment, you can use the BR tool to back up TiDB cluster data to S3-compatible storage, Google Cloud Storage (GCS) and persistent volumes (PV), and restore them:

• Back up Data to S3-Compatible Storage Using BR
• Restore Data from S3-Compatible Storage Using BR
• Back up Data to GCS Using BR
• Restore Data from GCS Using BR
• Back up Data to PV Using BR
• Restore Data from PV Using BR

Other documents about BR

• Use BR Command-line
• BR Use Cases
• BR FAQ
• External Storages