[[["容易理解","easyToUnderstand","thumb-up"],["確實解決了我的問題","solvedMyProblem","thumb-up"],["其他","otherUp","thumb-up"]],[["難以理解","hardToUnderstand","thumb-down"],["資訊或程式碼範例有誤","incorrectInformationOrSampleCode","thumb-down"],["缺少我需要的資訊/範例","missingTheInformationSamplesINeed","thumb-down"],["翻譯問題","translationIssue","thumb-down"],["其他","otherDown","thumb-down"]],["上次更新時間:2025-09-05 (世界標準時間)。"],[],[],null,["# Disaster recovery overview\n\nThis document provides information about disaster recovery tools and\ntechniques that help you protect your data from loss and downtime.\n\nSpanner disaster recovery features\n----------------------------------\n\nSpanner is designed to be scalable and globally distributed.\nSpanner offers the following features to help ensure high\ndata availability:\n\n- **Multi-regional configurations**: Spanner can maintain\n replicas of your data in separate zones\n within a single region or multiple regions to ensure data availability\n even if a zone or region fails.\n\n- **Database deletion protection**: You can protect against the\n accidental deletion of existing databases by users or\n service accounts that have the necessary Identity and Access Management (IAM)\n permissions to delete the database.\n\n- **Database back up and restore**:\n You can create backups of\n Spanner databases and restore them to help protect\n against operator and application errors.\n All backups are highly available, encrypted, and can\n be retained for up to a year from the time they are created.\n You can create full backups on demand or by using a backup schedule.\n You can only create incremental backups by using a backup schedule.\n\n- **Export and import**: You can\n export Spanner databases to\n Cloud Storage in either CSV or Avro format.\n\n- **Point-in-time recovery (PITR)**:\n Spanner point-in-time\n recovery (PITR) provides protection against\n logical data corruption, and accidental database deletion or writes. For example,\n if an operator inadvertently writes data or an application launch corrupts the\n database, you can recover the data from a point in time in the\n past (up to 7 days).\n\n- **Cross-region copy**:\n You can copy backups to different geographic\n regions to protect your data against regional failures or to meet your\n organization's compliance requirements.\n\n### Choose between backing up or exporting your database\n\nCompare backing up your databases to exporting your databases\nbefore deciding which one to use. For example, backups can have a maximum\nretention period of one year. You can consider exporting your database\nif you want a longer retention period than\none year. The following table describes similarities and differences between\nusing back up and restore versus using import and export:\n\nDisaster recovery techniques\n----------------------------\n\nSpanner offers disaster recovery techniques to protect\nyour data against the following disasters:\n\n- **Zonal failure**: Spanner regional configuration provides automatic protection against zonal failures, ensuring that if one zone within a region fails, your application remains operational.\n- **Regional failure** : If your application requires even higher data availability and protection against regional failures, use [multi-region configurations](/spanner/docs/instance-configurations) that offer 99.999% availability.\n- **Major geographical disasters** : Use Spanner [cross-region backup copy](/spanner/docs/backup#how-backup-copy-works) to make backups available in different regions.\n- **Logical corruption** : Use the following disaster recovery techniques for different retention periods:\n - Set up [point-in-time recovery (PITR)](/spanner/docs/pitr) to restore your data at any point in time within the last 7 days.\n - Set up a [backup schedule](/spanner/docs/backup/create-manage-backup-schedules) that creates full backups or incremental backups at a frequency that meets your requirements. All backups can be retained for up to one year.\n - [Export your database](/spanner/docs/import-export-overview) to Cloud Storage where you can retain the data for compliance, analytics, or reporting.\n- **Accidental database deletion** : Use [database deletion protection](/spanner/docs/prevent-database-deletion) to prevent the accidental deletion of existing databases by users or service accounts that have the necessary IAM permissions.\n\nOptimize costs for your disaster recovery strategy\n--------------------------------------------------\n\nYou can optimize Spanner disaster recovery\ncosts in the following ways:\n\n- **Multi-region configurations**: Use multi-region configurations only for applications that require 99.999% availability. For applications that require read-only latencies, consider read replicas in other regions.\n- **Backup frequency**: Schedule backups only as frequently as needed to meet your requirements.\n- **Backup type**: Use incremental backup schedules to save storage costs.\n- **Backup retention**: Determine and set the backup retention period for the shortest time required to meet your recovery and compliance needs.\n- **Large data exports** : Consider using Spanner [Data Boost](/spanner/docs/databoost/databoost-overview) for large data exports to offload compute resources from your instance and avoid negative impact on your transaction performance.\n\nTest your disaster recovery strategy\n------------------------------------\n\nConsider testing and validating the following components of your disaster\nrecovery plan:\n\n- Simulate the most likely events that can cause data loss for your organization.\n- Practice restoring your databases from a created backup. For more information on restoring your database, see [Restore overview](/spanner/docs/backup/restore-backup-overview).\n- Assess the impact of your disaster recovery plan on storage utilization.\n- Assess the impact of backup processes on your application performance.\n- Simulate a zonal or regional failure to test your failover and recovery procedure.\n\nWhat's next\n-----------\n\n- [Backups overview](/spanner/docs/backup)\n- [Create backups](/spanner/docs/backup/create-backups)\n- [Create and manage backup schedules](/spanner/docs/backup/create-manage-backup-schedules)\n- [PITR overview](/spanner/docs/pitr)"]]