Google Cloud Spanner provides externally consistent transactions, but its placement, recovery options, and costs depend on how you configure it. A regional instance keeps replicas in one region; dual-region and multi-region configurations distribute them geographically and are Enterprise Plus capabilities. Backups are point-in-time snapshots—not live replicas—and a sound regional recovery plan needs a backup copy outside the region it is meant to protect.
Is Cloud Spanner strongly consistent?
Yes. Google describes Spanner transactions as externally consistent: they are serializable, and the database’s transaction order matches the order clients observe commits. Google Cloud’s Backups overview states: “Backups of a Spanner database are transactionally and externally consistent at the versionTime of the backup.” That means a backup represents a consistent database snapshot at its recorded point in time, not changes made afterward.
How do Spanner regions and configurations work?
A configuration determines where replicas live and shapes the balance among data residency, user read latency, write latency, regional failure tolerance, and cost. Regional placement is a natural baseline when users and workloads are concentrated in one region. Geographically distributed configurations may help serve users closer to their locations and provide documented regional-failure resilience, but can cost more and affect write latency.
| Configuration | Placement and replicas | What to weigh |
|---|---|---|
| Regional | Three read-write replicas in separate zones within one Google Cloud region. | Suitable when workload geography and residency requirements center on one region. It does not provide the same cross-geography resilience as a distributed configuration. |
| Dual-region | Replicas are distributed across two regions; exact topology depends on the selected configuration. | Consider geographic resilience and read locality against write latency, pricing, and residency requirements. |
| Multi-region | Replicas are distributed across multiple regions; exact topology depends on the selected configuration. | Can support geographically distributed access and documented regional-failure resilience, with higher cost and possible write-latency trade-offs. |
Google lists dual-region and multi-region configurations as Enterprise Plus capabilities. Check the current Spanner editions documentation before choosing an edition or relying on a feature. Compare configurations for your specific workload rather than treating one as universally best.
Do these 3 things before closing this tab:
1Fix the driver behind crashes, sound loss and screen glitches2Repair Windows errors before they cause bigger problems3Scan for outdated or missing drivers - takes under a minute#1 Best Overall
How do Spanner backups work?
A backup is a retained snapshot at its versionTime, not a live replica. It includes database data, schema information, and database options set with ALTER DATABASE SET OPTIONS. It does not include subsequent changes, IAM policies, or change-stream data records. Backups are encrypted and highly available, and can be retained for up to one year.
A backup initially resides in the same instance, region, and project as its source database. If recovery or compliance requires an independent location, copy a completed backup to another region or project. That copy is not an instantaneous failover path: Google states that copying cannot proceed if either the source or target region is down. Account for the time needed to complete the copy and restore it when planning regional disaster recovery.
Rank #2
Google documents full backup schedules at intervals of 12 hours or longer and incremental schedules at intervals of four hours or longer. New instances have default backup schedules that create daily full backups with seven-day retention; those schedules take 24 hours to become active. Verify current edition and schedule support in the backup documentation before relying on a particular schedule.
Can you restore a Spanner backup in another region?
Yes, but first copy the backup to the destination region or project. Restoring creates a new database from the snapshot; it does not overwrite the source database. A restore may have an initial restore phase followed by optimization, so provision enough target compute and storage for the restored database and its expected traffic.
Rank #3
Several items are not automatically restored and may need separate handling:
- IAM policies: reapply the required access policies to the new database.
- Change-stream records: the backup does not include these internal records; account for that in downstream change processing.
- Row-deletion TTL policies: recreate them where required.
- Manually created split points: recreate them if the restored workload needs them.
The restore operation itself has no charge, but the destination’s compute, storage, and subsequent service use are billable. See Google’s restore documentation for operational details.
How much do Spanner backups and configurations cost?
There is no single monthly price: the bill depends on edition and configuration, compute capacity, database storage, backup storage, replication, and network usage. Google’s pricing page displays starting amounts, including SSD storage at $0.10 per GB per month per replica and HDD storage at $0.02 per GB per month per replica; backup storage at $0.10 per GB per month for regional configurations and $0.30 per GB per month for dual-region and multi-region configurations; inter-region replication starting at $0.04 per GB; and inter-region egress starting at $0.01 per GB. These are page-displayed starting rates, not a workload quote. Check the current Spanner pricing page and calculator with your region, currency, edition, capacity, storage, and replication assumptions.
Backup storage billing starts when backup creation completes and continues until deletion. Each completed backup has a 24-hour minimum billing period, even if deleted earlier. Copies incur backup storage charges like originals, and cross-region copies can also incur data-transfer charges. Restore itself is free; the restored database’s resources and usage are not.
What is the difference between a Spanner backup and an export?
Both preserve a transactionally and externally consistent view of the database, but they serve different purposes. A backup uses Google’s encrypted format for restoring into Spanner and, according to Google’s comparison, does not draw on instance server resources. An export writes CSV or Avro files to Cloud Storage as a medium-priority task, making the data more portable to systems that support those formats.
| Consideration | Backup | Export |
|---|---|---|
| Primary purpose | Restore a Spanner database from a point-in-time snapshot. | Make data available in CSV or Avro for other systems and workflows. |
| Format and destination | Google’s encrypted backup format; can be copied across regions or projects. | CSV or Avro files in Cloud Storage. |
| Retention | Up to one year. | Controlled by Cloud Storage retention settings. |
| Operational considerations | Designed for Spanner recovery; backup storage and possible copy-transfer costs apply. | Runs as a medium-priority task; consider Cloud Storage, Dataflow, and transfer costs. |
Choose a backup when straightforward Spanner recovery is the goal. Choose an export when portability or open file formats matter more, and include the separate storage and processing costs in the plan. Google’s backup and export comparison describes the service differences.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
