|Good Morning!Welcome to issue number 107 of Last Week in AWS. Thanks to everyone who pitched in with feedback on the new look.I’ll be in Atlanta for the first time this week for the AWS summit!Is your company in Atlanta? Want to talk AWS billing? I’d love to come chat with your team for an hour about your bill and ways to make it stop hurting. Send me a note and let’s get something set up. (Spots are limited, since, you know, there’s only so much time in the day…)|
This week’s issue is sponsored in part by GoCD.
Your CI/CD pipeline is broken. How do I know? Either I’m Nostradamus, or this is a very common problem. This week’s issue is sponsored by GoCD, from ThoughtWorks. It’s free and open-source, integrates natively with the most popular cloud infrastructures, and has both a broad community and enterprise support options for those of you who’re into either end of that very broad spectrum. Tools don’t solve issues, but GoCD eases your continuous delivery pains. Thanks to GoCD for their support.
From the Community
|This is a relatively entry-level approach to CloudFront that nonetheless manages to hit three issues I, not exactly entry-level when it comes to CloudFront, tripped over and went sprawling last week.A great intro to Aurora Serverless. Fun fact: the author and I were on a panel together at re:Invent 2017. Manish is great…I’m very confused about how using Parquet on Athena could possibly save money–wait. I’m thinking about “parkour” again.This article on Amazon’s fraught relationship with open source includes a bunch of people I know and respect, at least two glaring factual inaccuracies I can’t get into, and a name collision that caused a near-spittake. Instead of staking out a position that’s guaranteed to annoy at least half of my readership regardless of which side I take, I instead use this article to make a bunch of references that maybe five people will get. This is my newsletter, my jokes are for me!A short Twitter thread about EC2 network bottlenecking.Apple’s AWS bill is apparently nuts until you remember that they’ve recently stated they’re spending $10 billion on new data centers, relegating their AWS bill to “oh, that’s cute” territory.Amazon has a mascot?! WHAT?! And how is Billie the Platypus so much better than this lumpy bozo?!|
This week’s issue is sponsored in part by N2WS.
As a Systems Engineer, you’ve got one impossible challenge: to protect the applications running on AWS with the same level of scrutiny as you had when they were running on-prem. (Oh…and you need to retain backups for 7 years, which becomes prohibitively expensive with the amount of data flowing through your enterprise workloads.) But allow me to magically wave my digital wand over your AWS cloud and erase all of those worries! That’s essentially what N2WS Backup & Recovery does for your AWS cloud. N2WS allows you to cycle backups through different storage tiers so you can migrate critical apps and achieve not just the same but even better SLAs, while keeping costs lower than on-premises solutions–try it free today!
If you’ve got an interesting job for this newsletter’s emminently employable subscribers, get in touch!
It can often feel like keeping up with AWS is like drinking from a firehose. Why not work on an actual firehose? Kinesis Firehose (and other Kinesis teams) are solving hard problems around ingesting massive quantities of data at huge scale. With roles in Seattle, Palo Alto, and Cambridge (the one in the U.K.; they were banned from the one in Boston due to their insistence on pronouncing MIT as “mitt”), they’re a fascinating team with wonderful challenges. For more information, check out their open roles. For a lot more information, implement Kinesis Data Streams in your environment.How would you like an opportunity to apply your AWS skills with a company solving real-world problems and improves people’s lives and their health? Novartis Institutes for BioMedical Research is hiring for a Principal Cloud Engineer in Cambridge, MA. It sounds like an incredible role: regularly work across teams cross-functionally helping other teams to implement industry best-practices in their work and keep the AWS-based platform running smoothly. Have a look at the job posting for more details.
This week’s issue is sponsored in part by GitPrime.
GitPrime’s new book draws together some of the most common software team dynamics, observed in working with hundreds of enterprise engineering organizations. Actionable insights to help you debut your development process with data. Get Your Copy.
|Amazon EKS Supports EC2 A1 Instances as a Public Preview – Add “processor architecture” to the list of “random reasons your applications might fail in Kubernetes.”Amazon EC2 T3a Instances Are Now Generally Available – A fairly quiet launch of a fascinating new instance type; if you’re not CPU bound on your t3 instances (and if you are please, please, please us another instance type) these AMD variants will knock about 10% off of your cost.Amazon Elasticsearch Service now offers improved performance at lower costs with C5, M5, and R5 instances – This is a super uplifting headline, and beats my “year-old instance families freaking finally added to a service that feels half-abandoned.”Amazon Kinesis Data Streams changes license for its consumer library to Apache License 2.0 – Open source folks and corporate compliance folks can find common ground and rejoice over this one.Amazon MSK expands its open preview into AP (Singapore) and AP (Sydney) AWS Regions – “How do we boost our number of annual feature releases for the re:Invent keynote slide? I’VE GOT IT! Regional expansion of services before they go generally available!”Announcing the AWS Asia Pacific (Hong Kong) Region – The Hong Kong region is now live. It’s also disabled by default, and the various APIs throw different errors across the board. If your tooling mysteriously broke last week, well–you now know why.Amazon RDS now supports per-second billing – An exciting enhancement that arrives only 63072000 seconds after it should have.AWS specifies the IP address ranges for Amazon DynamoDB endpoints – “I’m configuring my firewall to permit access to DynamoDB” feels like trying to install a warp drive in your Model T.AWS Systems Manager Now Supports Use of Parameter Store at Higher API Throughput – Most of you won’t care, but this feature release addresses an issue that’s brought me the closest to screaming at an AWS employee in recent memory. Thank you. I see you. You know who you are.AWS Systems Manager Parameter Store Introduces Advanced Parameters– The answer to “how do we stop losing money on people choosing Parameter Store over (the 40¢ per record) Secrets Manager” is apparently to take a book from Patrick McKenzie‘s playbook and charge more.DynamoDBMapper now supports Amazon DynamoDB transactional API calls – I spent thirty seconds convinced I’d failed you, but no–DynamoDBMapper isn’t a service I missed; it’s a library–apparently mostly in Java?Now you can tag Amazon DynamoDB tables when you create them – …making it far easier for your usage to be attributed back to you. Now your CFO can tag you–with a bat.|
|Introducing Ian Mckay’s Former2: it takes your existing cloud resources, and converts them into CloudFormation / Terraform / Troposphere. Y’know, that thing you’ve been wishing for for ages, but AWS themselves whine consistently about how hard it is to build? Yeah, Ian built it in his spare time.Seventeen-year-old Andrew Taylor has independently created and released AWS’s service naming algorithm.I’m annoyed that “cycling a bunch of EC2 instances in an ECS cluster” still apparently requires a third party tool. Am I missing something?|
… and that’s what happened Last Week in AWS