Google Cloud Introduces Bigtable Tiered Storage
Google Cloud Bigtable: Preview of Tiered Storage
Google Cloud has launched the preview of Bigtable tiered storage, enabling developers to manage both hot and cold data within a single Bigtable instance.
This new capability helps optimize storage costs while maintaining seamless access to all stored data.
---
Key Features of Tiered Storage
Age-Based Tiering Policy
- Minimum threshold: 30 days
- Data moves automatically between:
- SSD (hot) tier
- Infrequent access (cold) tier
- No manual exports required for seldom-accessed data.
Migration is determined solely by the timestamp of a cell:
- When a cell’s timestamp exceeds the configured threshold, it is moved from SSD to infrequent-access storage.
- Movement is independent of read frequency.
---
Benefits Highlighted by Google
Anton Gething (Senior Product Manager) and Derek Lee (Software Engineer) emphasize:
> Tiered storage works with Bigtable’s autoscaling to optimize your instance usage. Data in the infrequent access tier remains available via the same API as SSD storage.
Advantages:
- Reduced operational overhead
- Eliminated need for manual data migration
- Seamless integration with existing query workflows
---
Optimal Usage Recommendations
To fully leverage SSD performance:
- Use timestamp range filters to query data stored exclusively on SSD.
- For analytics and reporting:
- Query cold data via Bigtable SQL
- Build logical views for targeted historical data access without exposing full datasets.
---
Capacity & Cost Improvements
- Storage capacity boost: Tiered-storage nodes offer 540% more capacity than standard SSD nodes.
- Cold storage pricing: Up to 85% cheaper than SSD storage.
- Not available for Bigtable HDD instances.
- Limitations:
- Bigtable Data Boost not supported
- Hot backups not supported
> To move data back to SSD:
> - Increase tiering policy age threshold
> - Disable tiered storage
> - Rewrite data with a new timestamp and delete the old copy
---
Real-World Impact
Florin Lungu, Lead DevOps Engineer & VP at Deutsche Bank, remarks:
> Bigtable tiered storage offers a solution to manage data costs without sacrificing data.
> This could significantly impact how organizations optimize their data storage strategies.
---
Related Innovations
Earlier this year, Google introduced tiered storage for Spanner, the managed distributed SQL database (InfoQ news coverage).
---
Complementary Tools for AI & Content Workflows
For developers and content creators adopting multi-tier data strategies, platforms like AiToEarn官网 can extend value by providing:
- AI-assisted content generation
- Multi-platform publishing
- Monetization
- Integrated analytics & model ranking
Connected platforms include:
Douyin, Kwai, WeChat, Bilibili, Rednote (Xiaohongshu), Facebook, Instagram, LinkedIn, Threads, YouTube, Pinterest, and X (Twitter).
Explore more:
---
About Bigtable
Bigtable is a managed, low-latency NoSQL database compatible with Cassandra and HBase. It supports:
- Structured
- Semi-structured
- Unstructured data
Recommended use cases:
- Time-series data from sensors
- Manufacturing and automotive operations
- Large-scale operational and analytical workloads
---
Final Takeaway
For organizations managing massive historical datasets, tiered storage:
- Reduces costs
- Expands capacity
- Simplifies integration into analytics workflows
Combining this with AI-powered tools like AiToEarn官网 can create monetizable, cross-platform outputs from analytics results—streamlining both technical and business operations.
---
Would you like me to also create a visual diagram showing how data moves between SSD and cold storage tiers in Bigtable? That could make this content even easier to digest.
