Welcome to issue number 41 of Last Week in AWS.
The fallout from Spectre and Meltdown continues to cause strife throughout the AWS ecosystem– and beyond. CPU usage numbers are a good 30% higher in many cases, and that’s causing consternation for some workloads. Stay safe out there…
This week’s issue is sponsored by MongoDB.
MongoDB Atlas is a cloud-hosted MongoDB service engineered and run by the same team that builds the database. Deploy, operate, and scale a MongoDB database in just a few clicks. Start free today.
I wrote some thoughts around The Business Value of Kubernetes this past week, as I’ve started advising ReactiveOps. More to come on this front; feel free to hit reply and tell me what I got wrong. It's Kubernetes-- how hard could it be?!
Grab talks numbers about the Meltdown patch’s impact on their AWS environment.
This is fun– a comparison of Image Recognition APIs as applied to images of food. I didn’t realize that Hotdog Or Not could be made to work with various brands of hotdog!
Cloudonaut released an AWS Monitoring Primer this week– holy complexity. Every time I look into this space I get a headache.
Cloudonaut presents a handy EC2 Network Performance Cheat Sheet that breaks down a lot of things we generally just handwave over and hope nobody asks us about.
A handy guide to the living nightmare that is Cognito User Pool based authentication. I’ve been tilting at this windmill myself for entirely too long…A quick and opinionated guide on how to set up MFA for AWS comes our way from Fernando Miguel, everyone's favorite AWS security fanatic.
Monitoring Weekly will likely hate this idea: a website monitoring tool built entirely in Lambda. I love that it’s called Uptime Girl, but I've done enough Billy Joel parodies in the last few months to last me a while.
A walkthrough of improving operations on Kafka and ZooKeeper in AWS.
Adrian Horn gave a talk on 10 Lessons from 10 Years of AWS that inspired this blog post. Sadly, nobody’s ever asked for my snarky “top ten lessons learned” from working with AWS.
Simon Wardley sits down with A Cloud Guru and talks about his philosophy on serverless, containers, and why he’s bullish on Amazon.
A handy guide on using Lambda and Fargate to handle ong-running processes in a Serverless application. I’ve got a couple of Last Week in AWS problems that this may well solve for me.
There’s an official AWS Alexa Skill out in the world. “Alexa, ask AWS what’s new” gives a list of recent launches. This handy skill doubles as a status page– if Alexa fails to respond, us-east–1 is probably on fire again.
A scaling benchmark exercise between Azure Functions and AWS Lambda, written by an Azure advocate. Fascinating deep dive on the two most widely used serverless implementations.
It turns out that when AWS tries to hire entire teams of engineers away from startup customers, those startups begin to take a hard look at whether or not they want to continue using AWS. Werner Vogels promises to look into it.
The bits and bobs of upgrading to C5/M5 instances, with an emphasis on what has to change for your workloads to run there. Device naming and drivers are different for NVMe devices, enhanced networking changes things somewhat, and they’re limited to a few region– but they’re also a lot cheaper.
Migrating your data from DynamoDB to MongoDB Atlas | MongoDB - The always-entertaining @jaydestro talks about migrating from DynamoDB to MongoDB. I classify both into the category of “datastores that confuse me so I let competent people handle them instead.”
Amazon Org Structure – Stratelogical - An amusing look at the various organizational structures of Amazon’s various divisions.
Amazon RDS Read Replicas Now Support Multi-AZ Deployments - This has been on the wishlist for a long time– you can now have multi-AZ read replicas, thus helpfully avoiding the Bad AZ in us-east–1 (spoiler: it’s all of them).
Amazon Rekognition Now Available in AWS US-East (Ohio) Region - Rekognition is now in Ohio. The delay was due to figuring out how to keep the software from recognizing that it was in Ohio and shorting itself out.
AWS Glue Now Supports Scala in Addition to Python - Now working with AWS Glue can be ever closer to working with actual glue, provided you’re an unsupervised toddler. Spoiler: it’s probably best left as an outdoor activity.
Blu Genius for Mainframe to AWS Cloud-Native Transformation | AWS Partner Network (APN) Blog - I missed this one when it came out last year– and it’s too good to pass up. If you’re in need of COBOL-to-Cloud, take a look at Blu Genius, and a harder look at the life choices that brought you here, to this moment.
Validate Your IT Security Expertise with the New AWS Certified Security – Specialty Beta Exam | AWS Security Blog - After crashing and burning the first time along, AWS has relaunched their security certification in beta mode. Apparently the first version told everyone to leave their S3 buckets open to the world.
Ooh, a CLI for AWS Fargate. This makes deploying containers from something like Jenkins a lot more straightforward…
A chrome extension that lets you tab between different AWS accounts. I like this a lot!
aws s3 sync with automatic / intelligent CloudFront cache invalidations, and you’re pretty close to websync.
As we breathlessly await official runtime support for it, you can run Lambda functions written in COBOL with this shim.
…and that’s what happened Last Week in AWS.
I’m Corey Quinn. I help people significantly reduce and understand their AWS bills and speak broadly on the conference circuit. I advise companies doing interesting things in the cloud space, such as ReactiveOps.
If you’ve enjoyed reading this, tell your friends to sign up at lastweekinaws.com (or post a link in your company Slack team!) about it. As always, if you’ve seen a blog post, a tool, or anything else AWS related that you think the rest of the community should hear about, send them my way. You can either hit reply– or join the #lastweekinaws channel on the og-aws Slack team.
List archives are always available at https://snarkive.lastweekinaws.com/