Dec 02 - Dec 08 Updates

Last Week Cloud Platform Updates

December 09, 2024

Dec 02 - Dec 08 | AWS, Azure & GCP Updates

AWS has introduced Amazon Aurora DSQL, a serverless, distributed SQL database designed for active-active high availability. Aurora DSQL delivers virtually unlimited scalability, 99.99% single-Region and 99.999% multi-Region availability, and effortless scaling for reads, writes, compute, and storage—all without sharding or instance upgrades. PostgreSQL-compatible, Aurora DSQL enables developers to build resilient, always-on applications with strong consistency and fast distributed SQL reads and writes.

AWS announces the preview of Amazon Bedrock IDE, a collaborative environment integrated into Amazon SageMaker Unified Studio. This platform empowers developers to create and customize generative AI applications with ease. Leveraging Amazon Bedrock's foundation models, Bedrock IDE provides tools like Retrieval Augmented Generation (RAG), task automation Agents, and responsible AI Guardrails. This unified workspace simplifies generative AI development, enabling seamless collaboration, faster prototyping, and deployment of AI solutions aligned with business objectives.

Amazon DynamoDB introduces multi-Region strong consistency for global tables, now available in preview. This feature enables applications to always read the latest data from any region, ensuring seamless and consistent operation across multiple locations.

AWS announces Amazon S3 Metadata (Preview), enabling near real-time updates and queryable metadata for S3 data. It supports system-defined and custom metadata, automatically updating as data changes. Metadata is stored in S3 Tables, optimized for tabular data, and integrates with AWS Glue Data Catalog (Preview) for seamless analytics with AWS services. Additionally, it supports AI-generated content annotations through Amazon Bedrock.

AWS has launched new AI Service Cards for Amazon Nova Reel, Amazon Canvas, Amazon Nova Micro, Lite, Pro, Titan Image Generator, and Titan Text Embeddings. These cards enhance transparency by detailing intended use cases, limitations, responsible AI design choices, and performance best practices.

Focusing on fairness, privacy, safety, and governance, the cards help customers make informed decisions for deploying AI responsibly. AWS plans to refine and expand these resources with ongoing customer and community feedback.

Now you can use index tuning to analyze the most resource intensive set of queries in your workload and produce recommendations of indexes that, once created, improve the performance of those queries. Index tuning also recommends indexes you can drop if they are identified as duplicates or are considered irrelevant for lack of use.  

The new index tuning feature builds on query store in Azure Database for PostgreSQL - Flexible Server, which captures and retains a history of queries executed in your workload and the runtime statistics of those queries.

In early-December 2024, the following updates and enhancements were made to Azure SQL:   

  • Get enhanced monitoring now that the dynamic management function (DMF) displays the component causing throttling of the primary log generation rate in Azure SQL Database Hyperscale. 

  • Save 35% to 55% on compute costs with a one-year or three-year reservation for Azure SQL Database, now also available for zone-redundant General Purpose databases. 

We're thrilled to announce that you can now modify multiple performance management server parameters in Azure Database for PostgreSQL – Flexible Server. You can specify the value of hash_mem_multiplier, wal_sender_timeout, and wal_receiver_timeout. You can also modify log_min_duration_sample, log_parameter_max_length, log_parameter_max_length_on_error, log_statement_sample_rate, vacuum_buffer_usage_limit, vacuum_failsafe_age, vacuum_multixact_failsafe_age , log_connections, and log_disconnections.

With the general availability of high-availability (HA) health status monitoring for Azure Database for PostgreSQL – Flexible Server, you can now track and maintain the health of HA-enabled instances with ease. This feature offers real-time insights into the readiness of your primary and standby replicas, helping you proactively identify and address any issues impacting availability or failover capability. Based on the Azure Resource Health Check framework, this monitoring solution enables you to optimize uptime and minimize potential disruptions to your database services.

Cloud SQL for MySQL, PostgreSQL, and SQL Server now supports using the Network Connectivity Center hub to propagate Private Service Connect endpoints within a VPC network. This integration enables transitive access to these endpoints across other spoke VPC networks through the hub, simplifying network connectivity for distributed environments. Currently available in Preview.

Starting with GKE version 1.31.1-gke.2105000, you can configure custom compute classes to utilize Compute Engine reservations. Workloads using these classes automatically consume reservations during node creation, enabling centralized management of reservation consumption.

A predefined Identity and Access Management (IAM) role is now available to grant Spanner-specific permissions for querying a Spanner database using Data Boost. This simplifies access management for leveraging Data Boost capabilities.