Alibaba Cloud Elasticsearch automatically creates snapshots for Elasticsearch clusters. You can directly restore snapshots stored in Alibaba Cloud Elasticsearch, or save snapshots to Object Storage Service (OSS) and then restore data from the snapshots. Alibaba Cloud Elasticsearch also supports load balancing. You can use these features to ensure the high reliability of Alibaba Cloud Elasticsearch clusters.

Automatic snapshot creation and data restoration from the snapshots

  • Automatic snapshot creation

    Alibaba Cloud Elasticsearch supports automatic creation of snapshots. You can enable this feature on the Snapshots page in the Elasticsearch console. You can specify the time when daily snapshots are automatically created based on your business requirements. This feature facilitates disaster recovery. For more information, see Snapshots.

  • Data restoration from automatically created snapshots
    After snapshots are created, you can restore data from the specified snapshots based on your business requirements. For more information, see Auto snapshot guide.
    Note
    • Alibaba Cloud Elasticsearch stores only snapshots that were created in the last three days.
    • You can restore data from automatically created snapshots only to the Alibaba Cloud Elasticsearch cluster where the snapshots are created.

Manual snapshot creation and data restoration from the snapshots

  • Manual snapshot creation
    Alibaba Cloud Elasticsearch allows you to save snapshots on your Elasticsearch cluster to OSS. If you want to use OSS, you must activate OSS and create a bucket in the same region as your Alibaba Cloud Elasticsearch cluster. You can also call the snapshot operation to create snapshots for specified indexes. For more information, see Snapshots and restore.
    Note You can store the snapshots only in OSS buckets of the Standard storage class. Snapshots cannot be stored in OSS buckets of the Archive storage class.
  • Manual data restoration from snapshots
    Alibaba Cloud Elasticsearch allows you to call the restore operation to restore index data from a specified snapshot. This feature facilitates disaster recovery. For more information, see Snapshots and restore.
    Note
    • You can restore data from snapshots in an OSS bucket only to an Alibaba Cloud Elasticsearch cluster that is in the same region as the bucket.
    • By default, each Alibaba Cloud Elasticsearch data node can process 40 MiB of data. You can modify the max_restore_bytes_per_sec parameter to adjust the data processing capability of the data nodes. For more information, visit Snapshot and restore on the open-source Elasticsearch website.

Load balancing

Alibaba Cloud Elasticsearch supports load balancing. You can specify the public or internal endpoint of your Alibaba Cloud Elasticsearch cluster on your client for access. Your requests are evenly distributed to all data nodes of your Alibaba Cloud Elasticsearch cluster based on load balancing.
Notice Load balancing among these data nodes depends on the number and size of index shards. When you create indexes, you must consider the number and size of index shards. For more information, see Shard sizing.