AWS re:Invent 2024 - Databases
Back to all Re:Invent 2024
Table of Contents
Blockchain
 |
AWS re:Invent 2024 showcases two use cases for Amazon Bedrock Agents in blockchain analysis and decentralized finance assistance. The first use case demonstrates how Bedrock Agents can enable natural language queries on public blockchain data sets, while the second use case explores how Bedrock Agents can be leveraged as a DeFi assistant to conduct research, manage crypto wallets, and protect users from security threats.
|
Databases
 |
This session discusses how to optimize generative AI applications with durable semantic caching in Amazon MemoryDB. It showcases the benefits of using MemoryDB's vector similarity search and semantic caching to improve performance, reduce costs, and enhance the user experience for generative AI applications.
|
 |
The video discusses the launch of Amazon Aurora PostgreSQL Limitless Database, a managed horizontal scale-out solution that provides the power of sharding with the simplicity of a single database cluster. The system leverages serverless technology to automatically scale up and down, offering millions of write transactions per second with consistent performance.
|
 |
AWS introduces new public blockchain datasets for Aptos, Arbitrum, Base, Provenance, and XRPL, in addition to existing Bitcoin and Ethereum datasets. The session also showcases a Bedrock agent-powered text-to-SQL solution that leverages these datasets to provide insights through natural language queries.
|
 |
The video provides a deep dive into Amazon Neptune, a managed graph database service, and its recent innovations. It highlights advancements in areas such as data management, query usability, performance improvements, and the integration of graph technology with retrieval augmented generation (GraphRAG) applications.
|
 |
The video discusses GE Vernova's migration of 600 TB of data from Apache Cassandra to Amazon Keyspaces, a serverless database service. The speaker, Yogini Parkhi, shares the challenges, strategies, and key learnings from the six-month-long transformation project, highlighting the importance of collaboration, observability, and a lean approach to managing complex data migrations.
|
 |
This video provides an in-depth look at AWS ElastiCache Serverless, its architecture, and the underlying technologies that enable it to achieve high performance and scalability. The presenters cover caching strategies, consistency models, and best practices for using ElastiCache Serverless, showcasing how it can simplify cache management and deliver low-latency responses at scale.
|
 |
The video presents a deep dive into Amazon DynamoDB's Zero-ETL integrations, showcasing how AWS simplifies data replication and ingestion by providing fully managed data pipelines. The presentation highlights the benefits of Zero-ETL, including increased agility, efficiency, and centralized data access, while demonstrating the ease of setting up a DynamoDB to SageMaker Lakehouse integration through a step-by-step console walkthrough.
|
 |
The talk provides a deep dive into the architecture of Amazon Aurora DSQL, a scalable and serverless relational SQL database optimized for transactional workloads. The speaker discusses the key design decisions, such as the use of a log-based architecture, multiversioning concurrency control, and a disaggregated system design, which aim to achieve high scalability, availability, and consistency.
|
 |
This talk provides an insider's look into the architectural choices and design principles behind Amazon DynamoDB, a highly scalable and available NoSQL database service. The speakers discuss how DynamoDB was built to handle high-throughput, low-latency applications by partitioning data, leveraging a distributed multi-tenant architecture, and engineering for consistent performance and availability.
|
 |
This presentation explores tips and best practices to boost performance and reduce costs in Amazon Aurora and Amazon RDS. It covers key cost dimensions such as compute, storage, and backup, and showcases strategies to optimize price and performance using real-world examples from a fictional startup, AnyCompany.
|
 |
The session provides an overview of Amazon MemoryDB, a highly available and durable in-memory database, and its new multi-region capabilities. It discusses the key considerations for choosing a multi-region database, such as consistency, availability, and performance, and how MemoryDB addresses these requirements through its active-active, asynchronous replication architecture.
|
 |
This talk provides a deep dive into the capabilities of Amazon Aurora Serverless v2, showcasing its ability to automatically scale compute resources up and down based on workload demands, enabling cost-optimization and reduced operational complexity for database management. The speakers also discuss how Aurora Serverless v2 integrates with other AWS services like Global Database and Redshift Serverless to provide a seamless and scalable data platform for various use cases.
|
 |
The video discusses the introduction of multi-region strong consistency (MRSC) for Amazon DynamoDB global tables, a feature that provides higher data consistency and availability compared to the existing eventually consistent global tables. The presenters explain the differences in the replication engines, the behavior during failure scenarios, and the trade-offs between the two global table options to help customers choose the appropriate solution for their use cases.
|
 |
The session discusses how AWS's zero-ETL integrations allow customers to easily connect their Amazon Aurora and RDS data to Amazon Redshift for near real-time analytics, without the need for complex data pipelines. The presenters highlight the security, performance, and reliability features of the zero-ETL solution, as well as recent enhancements and future plans to expand the integration capabilities across various data sources.
|
 |
This talk provides an overview of Amazon Aurora PostgreSQL Limitless Database, a managed horizontal scale-out solution that allows organizations to scale their databases beyond the limits of a single database. The presentation covers the architecture, data distribution, transaction handling, and query execution capabilities of this feature, highlighting how it addresses common scaling challenges faced by database engineers.
|
 |
The speaker discusses the new Amazon Aurora DSQL (Distributed SQL) database offering, highlighting its key features like scalability, serverless architecture, multi-region active-active support, and Postgres compatibility. He also delves into architectural patterns, isolation and consistency models, and best practices for building applications on Aurora DSQL.
|
 |
The video discusses the architecture and engineering behind Amazon DynamoDB, a highly scalable and performant NoSQL database service. It covers topics such as the MemDS in-memory data structure, warm throughput for predictable capacity, and the new multi-region strong consistency feature for global tables.
|
 |
The video discusses the use of generative AI in AWS DMS Schema Conversion to improve the efficiency and accuracy of database migrations. It provides an overview of the process, the challenges faced, and the plans for future improvements to make database migrations more seamless.
|
 |
This talk provides an overview of Amazon MemoryDB, a fast and durable vector database, and how Fidelity Investments leveraged its vector search capabilities to address their use cases around information barriers, self-help incident management, and the need for low-latency, secure, and scalable vector storage. The talk highlights Fidelity's decision-making process, architectural considerations, and the benefits they achieved by choosing MemoryDB as their vector store.
|
 |
This talk discusses how to supercharge app intelligence using generative AI with Amazon DocumentDB. The speaker covers vector search, access patterns, and best practices for implementing vector search on DocumentDB, as well as demonstrating a chatbot application that leverages DocumentDB and generative AI.
|
 |
The video discusses advanced data modeling techniques for Amazon ElastiCache, including the use of Valkey, a high-performance key-value data store. It covers various use cases such as caching, session storage, real-time analytics, and geospatial capabilities, as well as best practices and operational considerations for working with ElastiCache and Valkey.
|
 |
The video provides a deep dive into Amazon Aurora, including its innovations, architecture, and new features like local write forwarding, global database, and Aurora Serverless. It also introduces the new Aurora DSQL, a distributed and scalable version of Aurora, and compares its key differences with the traditional Aurora Postgres.
|
 |
The session covered the different types of blockchain wallets, from custodial to non-custodial and smart wallets, and how they can be securely implemented using AWS services like KMS, Cloud HSM, and Nitro Enclaves. The presenters discussed the key features and benefits of Nitro Enclaves for advanced crypto wallets, including strong isolation, cryptographic attestation, and cost-efficiency.
|
 |
This session provides a deep dive into the latest innovations and features of Amazon DocumentDB, including IAM authentication, vector search, global cluster improvements, and performance and cost optimization techniques. The presenters share insights from their work with customers to develop these features and highlight how DocumentDB is designed to scale and adapt to the evolving needs of enterprise document database workloads.
|
 |
The talk covers design patterns for achieving high availability and disaster recovery with Amazon Aurora, including cross-region backups, global databases, and multi-region configurations to ensure global resilience. The speakers discuss various trade-offs and best practices for building highly available and globally distributed database systems on AWS.
|
 |
The video discusses the launch of Oracle Database@AWS, a new offering that allows customers to run Oracle Exadata infrastructure within AWS data centers. It highlights the benefits of this offering, such as enabling customers to lift and shift their Oracle Exadata workloads to the cloud, unify their data, and leverage AWS services and capabilities.
|
 |
The talk discusses best practices for querying vector data in PostgreSQL for generative AI applications, focusing on the pgvector extension and techniques like approximate nearest neighbor search, filtering, and storage optimization. The speaker provides in-depth insights and guidance on making informed decisions about index building, data ingestion, and leveraging Amazon Aurora features to build efficient and high-performing vector search-powered applications.
|
Developer Experience
 |
This session discusses how Intuit, a leading fintech company, modernized its monolithic QuickBooks application by transitioning to a service-based architecture using cloud-native open-source databases like Amazon Aurora. The presenters share their journey, key decisions, and techniques used to achieve this transformation while ensuring minimal disruption to their 10 million customers.
|
 |
This talk provides an overview of AWS's perspective on Web3 and the various building blocks and services they offer to help customers accelerate their Web3 initiatives. The presentation also features a customer case study from Crypto.com, showcasing how they leveraged AWS services to build a world-class cryptocurrency exchange platform with exceptional performance, scalability, and stability.
|
 |
The presenters discuss core data modeling concepts for Amazon DynamoDB, including partition keys, sort keys, local secondary indexes, and global secondary indexes. They also cover techniques to optimize DynamoDB usage, such as data compression, soft deletes, and accurate counting of database operations.
|
 |
This session covers the evolution of chatbots and how to build serverless chatbots using Amazon ElastiCache and Aurora PostgreSQL. The speaker discusses the importance of caching, vector stores, and Retrieval-Augmented Generation (RAG) in creating optimized and personalized travel assistant chatbots.
|
 |
Vanguard, a global asset and wealth manager, leveraged AWS to rebuild and modernize its mission-critical trading platform, overcoming challenges of legacy infrastructure and increasing trading volumes. The modernization journey involved incremental delivery of value, leveraging cloud services like Aurora Postgres, DynamoDB, and Kafka, and achieving improved scalability, resiliency, agility, customer satisfaction, and reduced total cost of ownership.
|
 |
This talk covers advanced data modeling techniques for Amazon DynamoDB, a fully managed NoSQL database service. The speaker discusses key characteristics of DynamoDB, such as its consistent performance at any scale and Serverless-friendly nature, and provides practical guidance on data modeling, including the use of secondary indexes, DynamoDB Streams, and napkin math for cost and performance estimation.
|
 |
This talk provides a comprehensive guide on using AWS services and techniques to build data-driven generative AI applications. The key takeaways are to focus on the data workflow and leverage existing data sources, while automating as much of the process as possible to deliver personalized experiences.
|