DWS Snapshot Overview

Function

A snapshot is a backup of all data or incremental data (data that has changed since the last snapshot was created) in a DWS cluster at a specific point in time. It records the data in the current database and information about the cluster, including the number of nodes, node specifications, and database administrator username.

You can use snapshots to back up (manually or automatically) and restore clusters as needed. Backing up clusters is essential for maintaining data reliability, especially when the service provider cannot restore data through upstream re-import. This helps prevent data loss caused by humans or other factors.

• Backing up data using snapshots: Snapshots can be created manually or automatically.
  • Creating snapshots manually: You can manually create a snapshot to back up data before performing an important operation (upgrade, migration, or data change). For details, see Creating and Managing a DWS Snapshot.
  • Automated snapshot creation: You can configure an automated snapshot policy to enable the system to automatically create snapshots as scheduled. This method is suitable for routine data protection and does not require manual intervention. For details, see Configuring and Managing DWS Automated Snapshots.
• Restoring data using snapshots: You can restore data to the original cluster or a new cluster using snapshots. This is mainly used for data protection and fault recovery. For details, see Restoring a DWS Cluster.

Backup and Restoration Rates Using a Snapshot

The following table lists the backup and restoration rates using a snapshot. (They were obtained from the lab test environment with local SSDs as the backup media, and are for reference only. The rates may vary depending on the disk performance, network bandwidth, cluster load, and other factors.)

Snapshot Version Requirements

• Version requirements for table snapshots (If the function is unavailable on the console, contact technical support.)
  • To upgrade Roach 1.5.0 or later, contact technical support.
• Version requirements for schema snapshots (If the function is unavailable on the console, contact technical support.)
  • 9.1.0.100 or later
  • 8.3.0.110 or later 8.3.0.xxx cluster versions
  • 8.2.1.230 or later 8.2.1.2xx

Disk Space and Storage Space Constraints

• When the disk space usage exceeds 70%, data expansion and aging may cause backup failures. In this case, expand the disk capacity or add nodes in a timely manner.
• OBS snapshot storage space and billing
  • The cluster storage is provided by DWS free of charge. Cluster storage = Storage space per node x Number of nodes
  • DWS provides some free-of-charge storage space for you to store snapshot data generated in cluster backup. However, if you use more space than the free-of-charge storage space, the exceeded part is charged as per OBS billing rules. For details, see the OBS pricing details.

Constraints on Snapshot Dependencies and Deployment

• To create schema and table snapshots and use them for restoration, you need to authorize DWS to obtain database names, schema names, and table names. The obtained data will only be used for the fine-grained backup and restoration.
• The snapshot management function depends on OBS or NFS.
• If the backup device is an NFS backup medium, the NFS backup media must be mounted to the high-performance SFS Turbo. For details, see 11.1.3.2 Automatic Snapshot Policy.
• Only the snapshots stored in OBS can be used to restore data to a new cluster.
• If the default public schema is deleted, snapshot creation will fail.

Constraints on Cluster Creation from a Snapshot

• The new DWS cluster created based on the snapshot must have the same configurations as the original cluster. That is, the number and specifications of nodes, memory, and disks in the new cluster must be the same as those in the original cluster.
• If you create a new cluster based on a snapshot without modifying parameters, the parameters of the new cluster will be the same as those of the snapshot.

Constraints on Snapshots of Clusters with Decoupled Storage and Compute

• Only Storage-Compute Decoupled clusters running version 9.0.2 or later support snapshots. However, Storage-Compute Decoupled clusters with elastic virtual warehouses do not support backup or restoration. If a cluster contains elastic virtual warehouses, backup and restoration will fail. You are advised to reclaim the elastic virtual warehouses before performing backup and restoration.
• A snapshot created from a storage-compute decoupled cluster that uses hybrid billing can only be used to restore data to a yearly/monthly or pay-per-use cluster. For example, if a cluster has three nodes billed on a yearly/monthly basis and three nodes billed on a pay-per-use basis, restoring the cluster will change the billing mode of all six nodes to either pay-per-use or yearly/monthly.

Impact of Snapshots on Cluster Performance

• During snapshot creation, do not perform a VACUUM FULL operation, or the cluster may become read-only, affecting services.
• Snapshot creation affects disk I/O performance. You are advised to create snapshots during off-peak hours.
• During the snapshot creation, some intermediate files are retained, which occupy extra disk space. Therefore, create snapshots in off-peak hours and ensure that the disk capacity usage is less than 70%.
• Backing up multiple clusters with large amounts of data under the same account simultaneously may trigger OBS rate limiting, causing snapshot creation to fail.

Snapshots for Clusters of Different Types

Schema snapshots cannot be used for restoration. Exercise caution when creating them. To restore them, contact technical support.