Dec 4th, ‘24/4 min read
AWS re:Invent 2024 Day 2 Recap
Catch up on the highlights from AWS re:Invent 2024 Day 2, featuring key insights, exciting announcements, and key takeaways.
In case you missed it, here’s a recap of Day 1 at AWS re:Invent 2024!
I’ve compiled some of the key highlights and announcements from the day.
Highlights from Day 2
Here are some major announcements from the AWS keynotes by Matt Garman and Andy Jassy that caught my attention.
AWS Nova Foundation Models
AWS just launched Nova, which comes with a few exciting features:
- Micro: A text-only model, plus three multimodal options—Light, Pro, and Premier.
- Nova Canvas: For generating images, and Nova Reel for creating videos.
What’s really impressive about Nova is its 75% cost reduction and incredibly low latency compared to competitors.
Plus, its integration with Bedrock means you can fine-tune and add retrieval-augmented generation (RAG) features, making it an even more powerful tool.
New Trn2 Instances
- Powered by AWS’s latest Trainium2 AI chip, these instances offer 30-40% better price performance than the existing GPU-based EC2 instances.
- Amazon Q Business Intelligence Integration
The new Amazon Q Business features announced today enhance its strengths—helping employees find information and work more efficiently with their preferred tools while laying the groundwork for smarter generative AI experiences. These tools are built to save developers and data scientists time on routine tasks, letting them focus on the bigger picture.
Aurora DSQL:
A serverless, distributed SQL database that provides 4x faster reads and writes compared to alternatives, with strong consistency and high availability across regions.
The new capabilities are available to customers using either SQL (for storing and processing data in relational databases) or NoSQL (for non-relational databases that store data in a non-tabular format).
Next-generation AWS SageMaker
AWS SageMaker just got an upgrade too, and it’s now a unified platform for data, analytics, and AI development.
You can handle everything from data processing and search analytics to AI model development and even generative AI, all in one place. With its new unified studio, it’s designed to streamline the entire process of building and deploying AI solutions from start to finish.
I’m really excited to see how these innovations play out throughout the week!
Here are some talks I attended and found particularly interesting:
Observability the open source way
Cody, Marc, and Rodrigue shared their insights on using open-source tools for observability, and how to make the most of them without getting overwhelmed by their rapid development.
They showed how AWS-managed versions of popular tools like Prometheus, Grafana, and OpenTelemetry can help simplify the process, making it easier to standardize things like instrumentation, data collection, and analysis.
At Last9, we’ve taken the best of both worlds—open-source tools and managed solutions—without the usual headaches.
We've created a scalable telemetry data warehouse, fully compatible with Prometheus, designed to handle logs, traces, metrics, and events. It's already being tested by some large organizations.
Do give it a try if you haven’t!
Simplify, deploy, and scale an event platform: A community case study
Paloma and Ana shared how AWS built a website for a community event with a straightforward approach that doesn’t require deep cloud service knowledge.
They covered setting up user registration, call for papers (CfP), and managing speakers and sessions.
The session also showcased how open-source infrastructure as code tools can simplify deployment. It’s a session worth checking out once the recordings are available!
Use generative AI to optimize cloud operations for Microsoft workloads
Siavash, Ravindra, and Sassan shared how generative AI tools like Amazon Q and Amazon Bedrock can make managing Microsoft workloads in the cloud way easier.
They showed how one can use natural language queries to handle things like log management, configuration drift, and automation scripts—tasks that usually feel pretty daunting.
Accelerate innovation with AI-powered operations
Jared, Mihir, and Nikhil shared how AIOps can change the way you work, helping you spend less time putting out fires and more time being creative.
They walked through AWS best practices, especially using Amazon CloudWatch, showing how to tap into powerful tools that streamline operations and free up time for innovation.
The session was all about embracing the future of operations and unlocking your team’s full potential.
If you haven't met us at AWS re:Invent, we're here throughout the conference at Booth 166 with some fun stickers to make observability a little more relatable to you. Come find us!
Day 2 was packed with exciting AWS announcements, amazing talks, and some great discussions.
If you missed my Day 1 recap on LinkedIn, here’s my take: Folks are paying mansion prices for their observability studio apartment!
Prathamesh Sonpatki
Can’t wait to see what the rest of the event has in store!
Contents
Newsletter
Stay updated on the latest from Last9.
Handcrafted Related Posts
Challenges with Running Prometheus at Scale
Understanding limitations and challenges scaling Prometheus in modern cloud-native environments. Here we delve into long-term retention, downsampling, high availability, and other challenges.
Last9
Log Analytics 101: Everything You Need to Know
Get a clear understanding of log analytics—what it is, why it matters, and how it helps you keep your systems running efficiently by analyzing key data from your infrastructure.
Prathamesh Sonpatki, Anjali Udasi
Advanced OpenTelemetry: Sampling, Filtering, and Enrichment
OpenTelemetry offers powerful data collection, but maximizing its efficiency requires careful configuration. This article explores advanced techniques for sampling filtering, and data enrichment.
Anjali Udasi