#Amazon #Bedrock #Amazon #EC2 #Container #Service #Amazon #Elastic #Kubernetes #Service #Amazon
Origin | Interest | Match

EC2 high-RAM u7i instances support higher EBS-optimized instance performance
https://aws.amazon.com/about-aws/whats-new/2025/07/amazon-ec2-high-memory-u7i-instances-higher-performance/
With the latest enhancements to the AWS Nitro System, U7i instances now support up to 560,000 IOPS and 100 Gbps of EBS-optimized bandwidth. We recommend using io2 Block Express volumes for maximum IOPS performance. This increased EBS performance enables faster database operations and reduced restart times for large in-memory databases.
#AWS #EC2
Customer: "We'd like to save some money on licensing costs and switch from #RHEL to #AL2023"
Me: "Cool. Lemme check some things…"
Me: (Looks at Compliance as Code project) "Looks like Amazon isn't participating in that project, so, all our ready-to-go automation that leverages that won't work. Don't worry, it's not necessarily fatal, their distro has oscap
content, so, I'll check that…"
Me: (fires up and al2023 #EC2 and sees what Amazon makes available) "Looks like Amazon only makes available one hardening profile:
Title: Standard System Security Profile for Amazon Linux 2023
I have no freaking clue what that maps to, as far as common frameworks. Normally, for your needs, I'd recommend STIG or CIS, possibly even PCI/DSS or HIPAA. However, neither Amazon's tooling nor the Compliance as Code project seems to have those for AL2023. By point of comparison, I've checked not just the RHEL distro — that you wanted to avoid licensing costs on — but all the license-cost free ones that I know of — Oracle Enterprise Linus, CentOS Stream, Alma Linux and Rocky Linux and they all support STIG, CIS, PCI/DSS, HIPAA and others (all the same "others" that Red Hat does). I can't promise you that your accreditors are going to accept whatever the hell Amazon's "standard" hardening is, nor can I guarantee that your third-party scanning tools are going to be happy. All I can sorta guarantee is that if you use _Amazon's_ tools to verify _Amazon's_ profiles, that you'll probably get all greens if you stay up to date on patches. But you've also told me that you want to be multi-cloud by the end of next year: even ignoring your auditors and your thirde-party scanning tools, I'm not sure how helpful that's going to be to you.Oops. After I _did_ set up AWS budgets for my free-tier oriented accounts, I got sidetracked after only changing the credit_specification for the one that caused the problem in the first place.
Good to know those budget alerts work, I guess.
5-10m later and some tf validate/plan/apply runs later, and that's all sorted.
Note to self - When running a free tier EC2 instance (in a fresh AWS account) you can still end up being charged for that thing if you're not careful, when you overload it and run out of CPU credits.
At that point, if your “credit specification" has “Unlimited mode" enabled (and that appears to be the default) it will end up costing you for additional CPU credits.
Disabling that with TF, in the aws_instance:
credit_specification {
cpu_credits = "standard"
}
EBS Provisioned Rate for Volume Initialization
https://aws.amazon.com/blogs/aws/accelerate-the-transfer-of-data-from-an-amazon-ebs-snapshot-to-a-new-ebs-volume/
You can specify this volume initialization rate at which the snapshot blocks are to be downloaded from Amazon S3 to the volume.
With specifying the volume initialization rate, you can create a fully performant volume in a predictable time. If you run utilities like fio/dd to expedite volume initialization, it will remove the operational burden of managing such scripts with the consistency and predictability.
#AWS #EC2 #EBS
From the digital archives: #AWS #EC2 IP ranges from 14 years ago.
Little bit smaller than today (https://ip-ranges.amazonaws.com/ip-ranges.json)
AWS Systems Manager launches just-in-time node access
https://aws.amazon.com/about-aws/whats-new/2025/04/aws-systems-manager-just-in-time-node-access/
Admins can enable dynamic, time-bound access to nodes through policy-based approvals, controlling who can access which nodes and when. The policies determine whether an operator is denied access, automatically approved, or must obtain human approval before remotely connecting to nodes. Administrators can also increase visibility into RDP sessions by recording session activity and storing recordings in S3.
#AWS #EC2
New AWS Developers podcast episode: Rick Ochs explains how AWS Compute Optimizer analyzes resource utilization to provide EC2, EBS, and Auto Scaling recommendations.
Listen in your podcast apps or here
https://developers.podcast.go-aws.com/web/episodes/162/index.html