Google Cloud Bigtable Cheat Sheet

• A fully managed NoSQL database service designed for large analytical and operational workloads and enables you to store terabytes or even petabytes of data.

Google Cloud Bigtable Features

• You can use Cloud BigTable to store and query time-series data.
• It is ideal for storing large amounts of single-keyed data.
• Scales seamlessly from thousands to millions of reads/writes per second.
• Resize your cluster nodes to adjust Cloud Bigtable throughput without restarting – all without downtime.
• SQL and continuous materialized views: Use familiar SQL syntax with Bigtable. Build incremental materialized views that automatically update in real-time without impacting write/read performance.
• Data model flexibility: Store scalars, JSON, Protocol Buffers, Avro, Arrow, embeddings, images. Dynamically add/remove columns as needed.

• Easy migration from NoSQL databases: Apache Cassandra and HBase APIs with no-downtime live migrations. Bigtable Data Bridge simplifies migrations from Amazon DynamoDB.
• Multi-region deployments: From single zone up to eight regions with automatic replication. 99.999% availability with multi-primary configurations.
• Data Boost: On-demand, isolated compute resources for batch processing, analytics, and ML training without affecting transactional workloads.
• Change streams: Real-time change data capture (CDC) for analytics, event triggering, and compliance.
• Enterprise-grade security: Customer-managed encryption keys (CMEK), VPC-SC, Access Transparency, Access Approval, fine-grained access control at table, column, or row level.
• Observability tools: Key Visualizer, query stats, table stats, hot tablets tool for troubleshooting performance issues.
• Backups: Instant, incremental backups. Restore between instances or projects. Store backups in different regions for disaster recovery.
• Vertex AI Vector Search integration: Index Bigtable data with Vertex AI for similarity search over vector embeddings.
• LangChain integration: Built-in kNN nearest neighbor vector search (Preview) for generative AI applications.

Pricing

Bigtable pricing follows a pay-as-you-go model with several components:

• Compute capacity: You pay for the nodes provisioned in your clusters. Costs scale with the number of nodes and runtime.
• Data Boost: Optional serverless compute for batch workloads, billed separately from regular node usage.
• Storage: You pay for the amount of data stored, with different rates for SSD (low-latency) and HDD. Each replica in a multi-region setup incurs separate storage charges.
• Backups: Incremental backups are billed based on the amount of data stored.
• Network: Inbound traffic is free. Outbound traffic within the same region is free. Cross-region egress and replication incur charges.

Committed use discounts are available for longer-term commitments to reduce costs. There are no IOPS charges, and no additional cost for taking or restoring backups.

For current pricing details, refer to the official Google Cloud Bigtable pricing page.

Google Cloud Bigtable Cheat Sheet References:

https://cloud.google.com/bigtable
https://cloud.google.com/bigtable/docs/overview