The Cloud Pod: Oh the Places You’ll Go at re:Invent 2021 — Episode 144

On The Cloud Pod this week, as a birthday present to Ryan, the team didn’t discuss his advanced age, and focused instead on their AWS re:Invent predictions. Also, the Google Cybersecurity Action Team launches a product, and Microsoft announces a new VM series in Azure.

A big thanks to this week’s sponsors:

• Foghorn Consulting, which provides full-stack cloud solutions with a focus on strategy, planning and execution for enterprises seeking to take advantage of the transformative capabilities of AWS, Google Cloud and Azure.

• JumpCloud, which offers a complete platform for identity, access, and device management — no matter where your users and devices are located. 

This week’s highlights

• AWS releases new G5 instances, which feature up to eight NVIDIA A10G Tensor Core GPUs. That’s super, super fast. 
• Google’s Cybersecurity Action Team adds Risk and Compliance as Code (rCaC) Solution.
• Microsoft announces the NDm A100 v4 Series, and claims another spot on the  TOP500 supercomputers list.

Top Quotes  

•   “[AWS Resilience Hub] is already building on top of the FIS, which is interesting, but at some level I just want you to execute Lambda functions that validate things for me, and then tell me that I’m resilient because I validated it with Lambda.”
• “Anything that empowers more dynamic and interactive web development I’m all for.”

Amazon Web Services: Give Us Your Car  

• AWS is releasing new G5 instances, which feature up to eight NVIDIA A10G Tensor Core GPUs. For the cost of a small car every month, you too can get up to 40% better value on inferencing and graphics-intensive operations.
• AWS is releasing the Resilience Hub, a service designed to help you define, track and manage the resilience of your applications.
• Unified Search in the AWS Management Console now sources results from blogs, knowledge articles, events and tutorials. Buyer beware with this one: It will pull outdated information that is still available on AWS, and you could end up with a giant albatross that costs you a fortune.
• Amazon ECS is improving ECS Capacity Providers to deliver faster cluster auto scaling. When you’re using a capacity provider, it’s painfully slow to get the underlying hosting infrastructure to scale fast enough, so we’re presuming AWS has addressed this in the back end.
• Manage access centrally for JumpCloud users with AWS Single Sign-On. We’re super happy to see this: Take notes, Azure AD.
• Amazon ECS adds container instance health information. This is nice to see and will help improve your application resiliency.

AWS re:Invent 2021 Predictions    

• Prediction rule: If it’s already been officially announced by Amazon, then it doesn’t count. It needs to be in the rumor mill and somewhat specific.
• Each contestant will also predict how many total announcements are made, which will be used in the case of a tiebreaker.
• Jonathan

1. New silicon from Annapurna Labs, for a new Inferentia chip
2. AWS will partner with data centers in different countries, especially in Europe
3. More customer insights into their carbon footprints

• Tiebreaker: 21
• Ryan

1. A dashboard or visualization to help you manage organization-level networking
2. Data tokenization, anonymization and/or data scrubbing as a service or DLP solution
3. App configuration on demand

• Tiebreaker: 10
• Justin

1. Graviton 3, or whatever AWS names its new version of a general purpose ARM-based processor
2. Zero Trust networking solution
3. New Honeycomb features, and a lecture on the benefits of no-code

• Tiebreaker: 20
• Peter

1. Babelfish for Oracle
2. Serverless Neptune
3. Serverless DocumentDB with MongoDB compatibility

• Tiebreaker: 4

Google Cloud Platform: A Bit of a Stretch

• Memorystore for Redis Read Replicas is now in public preview, supposedly allowing users to scale their applications’ read requests by 6x. This is a bit of a stretch: The read requests are not simultaneously going to all six replicas.
• The Google Cybersecurity Action Team announces the launch of Risk and Compliance as Code (rCaC) Solution. While easily mistaken for a terrible action movie sequence, this is actually pretty cool because it’s built into the platform, so there’s no circumventing it. 
• Google announces the general availability of Vertex Pipelines. Vertex Pipelines is the best way to scale your machine learning (ML) workflow and run it as a pipeline, where each pipeline step is a distinct piece of your ML process. 

Azure: Thirsty Work

• Microsoft announces the availability of the NDm A100 v4 Series, featuring NVIDIA A100 Tensor Core 80 GB GPUs. Microsoft has now claimed four spots on the TOP500 supercomputers list. We don’t know the price tag but it’s probably expensive.   
• Azure Archive Storage rehydration priority update is now generally available. Look at Azure coming up with more ways of charging money for dumb features. 
• Azure Web PubSub is now generally available. Not to be confused with Google’s PubSub product.
• ✈️ JetStream Disaster Recovery for Azure VMware Solution is now generally available. You can now pay for a capability you can get from most cloud providers for free.

TCP Lightning Round

⚡ Jonathan elbows the others out of the way with a zinger at the end and takes the point this week, leaving the scores at Justin (16), Ryan (12), Jonathan (14), Peter (2). 

Other Headlines Mentioned:

• Amazon Kendra releases SharePoint Connector to enable SharePoint site search
• Azure Database for MySQL – Flexible Server is now generally available
• Public preview: Run Command – Execute PowerShell or shell scripts on Virtual Machines and Scale Sets for Azure
• New and improved Amazon Athena console is now generally available
• AWS Application Migration Service now supports agentless replication 
• Auto-instrumenting a Python application with an AWS Distro for OpenTelemetry Lambda layer
• Amazon Rekognition improves accuracy of content moderation for images
• SQL Server on Azure Virtual Machines: Multi subnet high availability generally available