Industry News – AWS re:Invent Week One Recap
Today wraps up the first week of AWS’ three-week virtual re:Invent conference, which Amazon describes as the “biggest cloud computing event of the year”—and the cloud provider did not disappoint. AWS unveiled a variety of announcements, innovations and strategies that are making it easier and more cost-effective for organizations to migrate and optimize their businesses on AWS. Here are our takeaways from week one.
New Instances Deliver Faster Networking and CPUs
Every year at re:Invent, AWS announces new instances and upgrades to existing instances. So, while it’s not surprising that week one of the conference followed this pattern, it’s a testament to AWS’ continued commitment to improving its underlying cloud infrastructure and providing customers with higher performance at a lower price tag.
Here’s a breakdown of the three big instance-focused announcements:
- New AWS Graviton2 Arm-based instances provide faster network bandwidth and higher performance at a more cost-effective price. Specifically, the new instances will “deliver up to 100 Gbps network bandwidth, up to 38 Gbps Amazon Elastic Block Store (EBS) bandwidth, up to 40% higher packet processing performance, and up to 40% better price/performance versus comparable current generation x86-based network optimized instances.”
- AWS released M5zn instances in seven sizes, all of which use a new generation of Intel Xeon Scalable processors. This powerful combination offers higher memory instances with faster access to storage—which is especially helpful for database servers.
- Amazon also unveiled a new Mac instance, which makes it possible to run a Mac on AWS. According to Amazon’s blog on the news, developers can “use Amazon EC2 Mac instances to build, test, package and sign Xcode applications for the Apple platform including macOS, iOS, iPadOS, tvOS, watchOS, and Safari.” This announcement is especially relevant to developers that want to develop Mac software, but lack access to Macs in-house, as well as individuals who use creative software that is only available on a Mac.
Serverless Computing Becomes More Accessible and Cost-Effective
Serverless computing is a hot technology right now, so it makes sense that AWS would make serverless product announcements a key focus at re:Invent—here are four that will accelerate its adoption and make it easier to manage and more cost-effective for developers.
- AWS Lambda billing increments decrease from 100 milliseconds (ms) to 1 ms. One of the greatest benefits of serverless computing is that costs are based on actual usage. Prior to this week, AWS Lambda users were charged for the number of times their code was triggered along with the amount of time their code executes, which was rounded up to the nearest 100 ms. Now, AWS is “rounding up duration to the nearest millisecond with no minimum execution time.” Going from 100 ms to 1 ms billing increments, means the cost per execution aligns much more closely to your actual usage—making it even cheaper to run your code in Lambda.
- Developers now benefit from increased memory and vCPUs available for Lambda functions. According to AWS, developers can now dedicate up to 10 GB of memory to a Lambda function (a 3x increase compared to previous limits) and access up to 6 vCPUs in each execution environment. Related to this announcement, AWS also announced the ability to use customer container images of up to 10 GB in size in Lambda.
With these two developments, AWS has increased the sophistication of code that can be handled by Lambda, so developers can do more in the Lambda environment before they have to use separate containers for each function and a pool of virtual machines to run said containers. This makes serverless computing more accessible to developers who engage in complex application development that would typically require containers.
- AWS gave a preview of AWS Proton, a new tool that “automates and manages infrastructure provisioning and code deployments for serverless and container-based applications.” Proton will handle the creation and ongoing updates of all PaaS components used to deploy a serverless application—making serverless computing not only more accessible and more cost-effective for developers, but also easier to manage.
- AWS also introduced version 2 of Amazon Aurora Serverless in preview. Originally introduced at re:Invent 2017, this latest version features an improved response to scaling. Amazon says v2 “scales to hundreds of thousands of transactions in a fraction of a second, delivering up to 90% cost savings compared to provisioning for peak capacity.” With this capability, developers can use a familiar relational database in a serverless environment, without having to worry about scaling based on demand.
AWS Babelfish Eliminates Refactoring and Licensing Constraints During Migrations
There have been occasions recently when enterprise vendors have forced their customers to rebuy licenses when they want to move a database from on-premises to the cloud. Alternatively, if the customer wants to move off the proprietary database to an open source offering, they have to refactor for the cloud, which can be a costly and complex undertaking.
AWS is helping customers overcome these traditional barriers with Babelfish for Amazon Aurora. In technical terms, Babelfish “is a new translation layer for Amazon Aurora that enables Aurora to understand queries from applications written for Microsoft SQL Server. With Babelfish, applications currently running on SQL Server can now run directly on Aurora PostgreSQL with little to no code changes.” In simple terms, this means you can now migrate SQL Server databases to Aurora PostgreSQL without having to refactor for AWS or purchase more licenses.
AWS Puts a Focus on Storage Speed
Two storage-focused announcements caught our eye, both of which have an underlying theme of accelerating storage speed and slashing costs:
- AWS launched a new type of SSD EBS volume, gp3, which allows you to “provision performance independent of storage capacity, and offers a 20% lower price than existing gp2 volume types.” Gp3 volumes offer four times the throughput of gp2, with a 3,000 IOPS baseline performance. Meaning, you no longer need to create big disks to get the IOPS you need, and you benefit from faster storage performance at a reduced price.
- AWS previewed io2 Block Express EBS volumes, which deliver quadruple the IOPS and throughput of the io2 EBS volumes. Instances have faster access to storage, and you’re running a faster storage mechanism on a faster storage network. Specifically designed for performance- and IO-intensive applications and databases, Block Express EBS volumes give end users the confidence to scale large data sets easily, quickly and cost-effectively.
Hybrid Infrastructure Developments Support Niche Use Cases
AWS is a cloud provider, but the company understands that there are some use cases that require organizations to maintain a hybrid environment. For example, a manufacturing company requiring a low latency connection between its manufacturing units and the servers running them. Or a bank that wants to run customer-facing functions in a friendly cloud environment, but is required by compliance mandates to keep certain data on-premises.
To support these types of deployments, AWS issued two relevant announcements:
- There are now two new smaller AWS Outposts form factors available to “give customers access to AWS on-premises in space-constrained locations.” Amazon defines AWS Outposts as “a fully managed service that offers the same AWS infrastructure, AWS services, APIs and tools to virtually any datacenter, co-location space or on-premises facility for a truly consistent hybrid experience.” Prior to this, if you wanted to leverage the AWS Outposts service, you had to purchase a full rack of capacity. Now, you have the option to purchase only 1U and 2U servers—which provide the same AWS capabilities in a smaller form factor, which are more affordable and require significantly less power and network connectivity.
- The two ways to run containers on AWS—Amazon Elastic Container Service (ECS) and Amazon Elastic Kubernetes Service (EKS) —can now be extended to the data center. In other words, you can install ECS or EKS as a VM within your datacenter, and Amazon will manage the containers for you.
Stay tuned next Friday for our wrap-up of week two of AWS re:Invent! To learn more about how Navisite can help you migrate and optimize your business on AWS, learn more about our AWS Managed Services.