Good Morning!
Today I’m in Kansas City keynoting NANOG, while tomorrow I’m off to catch the second half of Monitorama in Portland. If you’re in either place, please say hello.
And if you’re in San Diego next week, I’m hosting a drinkup at PLACE, on DAY at TIME. Come let me buy you a drink and you can tell me exactly what my problem is in person. JANICE: NEED INFO ABOUT DRINKUP IN SAN DIEGO TO GO HERE.
From the Community
Windows Recall demands an extraordinary level of trust that Microsoft hasn’t earned says Ars Technica, and I agree. When even the legendary Charlie Stross referes to this as the product nobody wanted, I can’t quite fathom how the company thinks going from its recent security issues to demanding this level of customer comfort is going to play out to its benefit.
Things you wish you didn’t need to know about S3 is a great article; expect a Screaming in the Cloud podcast episode with its author to come out in a week or two delving into this.
Just a heads up; if you’re using Mac Bartender to manage your menu bar on MacOS, I strongly suggest you delete it immediately. It’s got new ownership who are either data thieves or else acting exactly like data thieves act.
Our Unusual Journey to ECS on EC2 is certainly unusual…
Jobs
fractional/part-time head of service delivery – empty
full-time – empty
1099 contractor/part-time – empty
Podcasts
Last Week In AWS: What I Want to Know About Matt Garman
Screaming in the Cloud: Building Smarter Authorization Systems with Sam Scott
Screaming in the Cloud: The Pros of On-Prem Kubernetes with Justin Garrison
Choice Cuts
Introducing Amazon EC2 High Memory U7i Instances – Yeah, the top end of this costs $407.68 an hour. Bring money, even by SAP HANA standards.
Amazon API Gateway integration timeout limit increase beyond 29 seconds – Well if this isn’t the end of an era, what is? This was a constant hard limit since the first days of API Gateway. I wonder what use cases this is going to empower suddenly?
Amazon CloudWatch GetMetricData API now supports AWS CloudTrail data event logging – You can finally get some insight into which of your 14 observability tools is driving up your AWS bill with redundant CloudWatch metric calls. Hint: it’s all of them.
Amazon CloudWatch Logs announces Live Tail streaming CLI support – That’s cool I guess; cw is open source and has done this for ages for those of us not willing to wait for AWS to eventually get around to this. As a bonus, unlike CloudWatch Live Tail it doesn’t cost anything extra.
Amazon CodeCatalyst now supports GitHub Cloud source code with blueprints – What the hell is GitHub Cloud? There’s GitHub Enterprise Cloud, but calling it "GitHub Cloud" makes it sound like an out of touch luddite asking about "The Internets." AWS is far too prickly about service naming ("No, no, it’s ‘Bottlerocket,’ not ‘Amazon Bottlerocket!’) to be playing this fast and loose with other companies’ product names.
Amazon EC2 instance type finder capability is generally available in AWS Console – It’s hidden in a tiny "Get Advice" tooltip in the Instance Wizard and its recommendations are highly suspect, but good work I guess.
AWS Transfer Family increases message size and throughput limits for AS2 – Excitingly you now get to not have to request a limit increase until you’re spending $51 million a month on this monstrosity.
Integrate AWS Cost Anomaly Detection Notifications with IT Service Management Workflow – Part1 Jira – This is neat, but real enterprises are holding out for the ServiceNow integration.
Introducing CloudFront Hosting Toolkit – I always wondered when AWS would take a whack at competing with Vercel, Render, etc. but hadn’t considered what form that competition would take. An open source toolkit to wrangle the existing bad-UX services in a usable way feels like the least possible investment in this problem that they felt they could plausibly ship.
… and that’s what happened Last Week in AWS.
