Google Cloud Introduces AlloyDB Compatible with PostgreSQL for Enterprise Database Workloads

Google Cloud recently announced AlloyDB for PostgreSQL, a PostgreSQL-compatible managed service targeting enterprise deployments. AlloyDB is a comprehensive cloud database supporting Atomicity, Consistency, Isolation, and Durability (ACID) compliant transactions.

AlloyDB is compatible with PostgreSQL 14, providing portability of existing workloads and transitioning out of legacy databases. Google claims it’s “four times faster for transactional workloads and up to 100 times faster for analytical queries” than a standard PostgreSQL deployment. Andi Gutmans, general manager and vice president of database engineering at Google Cloud, writes:

Like many managed database services, AlloyDB automatically manages database patches, backups, scaling, and replication for you. But it goes a step further by using adaptive algorithms and machine learning for PostgreSQL vacuum management, storage and memory management, data tiering, and analytics acceleration. It learns your workload and intelligently organizes your data into memory, super-fast secondary cache, and durable storage.

In a separate article “AlloyDB for PostgreSQL under the hood”, Ravi Murthy, Director of Engineering at Google, and Gurmeet Goindi, Director of Product Management at Google, explain how the storage layer works:

AlloyDB starts by separating the database layer from the storage, introducing a new intelligent storage service, optimized for PostgreSQL (…) The fully disaggregated architecture, even at the storage layer, allows it to function as an elastic, distributed cluster that can dynamically adapt to changing workloads, adds fault tolerance, increases availability, and enables cost-effective read pools that scale read throughput horizontally. Multiple layers of caching in the stack (…) provide developers with increased performance while maintaining the scale, economy, and availability of cloud-native storage.


Many experts suggest that AlloyDB is the answer to Amazon Aurora, with Google Cloud claiming that AlloyDB can be twice as fast for transactional workloads as the AWS service. Unlike Amazon Aurora, AlloyDB does not charge for I/O but does not currently offer a serverless version. Gutmans adds:

A few people have contacted me asking how the “disaggregated” compute and storage used in AlloyDB (and Google systems) is different from the compute storage separation that other vendors talk about (…) is quite different, as in our model both compute and storage are “extremely” flexible at the most granular level and with very low latency. Others tend to have high latency between compute and storage and/or co-location of the two.

Marc Callaghanpreviously Distinguished Engineer at Mongo and MTS at Facebook, comments:

Cloud-native PostgreSQL is there (Yugabyte, AlloyDB, Amazon Aurora), but upstream PostgreSQL is the only way to take advantage of fast storage devices (low latency + direct connect). Will be interesting to see if that changes.

In a popular Reddit thread, some users point out the lack of a serverless pricing model and the limited features compared to Cloud SQL for PostgreSQL. Last year, Google Cloud introduced a PostgreSQL interface for Spanner, as previously reported on InfoQ.

Compute resources for AlloyDB are priced per vCPU and GB of memory and storage prices vary by region. For networking, entry is free and exit pricing varies by destination.

Maria H. Underwood