Flexible Server, Microsoft’s fully-managed Postgres service built on open-source Postgres now incorporates the power of Azure OpenAI directly into the database.

Azure Database for PostgreSQL powered by the PostgreSQL comes in two deployment modes: Flexible Server and Single Server.

The Flexible Server is a fully managed database service designed to provide more granular control and flexibility over database management functions and configuration settings. The idea is that we get an AI boost with the new functionality in azure_ai Extension for Postgres, directly integrating the database with Azure OpenAI.

This extension allows you to invoke the Azure Open AI API in order to create embeddings and store them on PostgreSQL.
This in turn enables you to implement capabilities such as semantic search, recommendations, and anomaly detection faster and more easily than ever before, and without extensive application changes. As a bonus, another option offered by the new extension is the ability to invoke the Azure language services such as sentiment analysis right from within the database. Of course, for turning PostgreSQL into a vector store, there's need for another Postgres extension, that of pgvector.

pg_vector is an extension for PostgreSQL that renders it a viable alternative to specialized vector stores used in LLMs. It allows you to replace specialized vector engines with Postgres, so that you can have your embeddings stored alongside your JSON or relation data under the same roof. That capability renders Postgres usable in an AI or ML setting with the use of azure_ai. At the same time when being able of hosting embeddings and without even touching the LLM yet, we still can do useful similarity searches like KNN and ANN from within Postgres. The updated version of Flexible sever gets upgraded to pg_vector 0. 51.

But the new features are not constrained to AI additions. There's also other extensions released as GA: azure_storage & tds_fdw.

azure_storage is used to simplify loading and exporting data between Azure cloud storage and Postgres, while tds_fdw is used as a foreign data wrapper to query data stored in Microsoft SQL databases.

New extensions aside, there's also a set of features that provide advanced storage and compute with optimal price-performance
for enterprise production workloads. These features can also help organizations optimize resources and improve mission-critical workloads:

• Premium SSD v2 which offers sub-millisecond disk latencies for demanding IO-intensive workloads.
  
  
• Public preview of IOPS scaling, which enables you to scale up IOPS to perform transient operations, such as migrations or data loads, more quickly and then scale back down when not required to save cost.
  
  
• General availability of near-zero downtime compute/storage scaling, which allows you to dynamically adjust the amount of your compute and storage resources based on current demand via a seamless experience with near-zero downtime.

More second grade changes are:

• Public preview Multi-Region Disaster Recovery (GeoDR)
  
  
• Public preview Long-term Backup Retention
  
  
• Server Logs for Azure PostgreSQL Flexible Server.
  
  
• Support for TLS Version 1. 3 – Enhanced Connectivity Security.
  
  
• Microsoft Defender support

And we left the best for last; PostgreSQL 16 support. Needles to say, Flexible also gets all the goodies that come with version 16, such as :

• Query performance improvements with more parallelism
  
  
• Developer experience enhancements
  
  
• Monitoring of I/O stats using pg_stat_io view
  
  
• Enhanced security features
  
  
• Logical replication from read replica servers

More Information

Azure Database for PostgreSQL - Flexible Server
 

Related Articles

Azure CosmosDB for PostgreSQL Adopts Version 16

Turn PostgreSQL Into A Vector Store  

To be informed about new articles on I Programmer, sign up for our weekly newsletter, subscribe to the RSS feed and follow us on Twitter, Facebook or Linkedin.

Comments

Make a Comment or View Existing Comments Using Disqus

or email your comment to: comments@i-programmer.info