Posts

• Multitenant inference architecture with SageMaker endpoints

  With Michael Pelts, Senior Solutions Architect at AWS and I explored ways to allow tenants to expose their own models for inference when ensuring that it is accessed only by the tenant that created it.

  We gave the talk online at the Australia Multicloud Engineering Meetup, then again face-to-face at the Tel Aviv location.

  “Multitenant inference architecture with SageMaker endpoints”.

  This talk has similar content to the article “Implementing SaaS Tenant Isolation Using Amazon SageMaker Endpoints and IAM” but also gives a higher-level architecture perspective.

• Throughput Metrics Across the Clouds

  I wrote an open-source project measuring throughput (also latency) as compared to distance between cloud regions, across clouds (GCP and AWS) or within a single cloud. From what I can tell, this is the first tool

  • measuring throughput (most others focus on latency)
  • and doing so across clouds
  • and relating it to distance.

  This article describes the conclusions.

• Re:Invent 2022: What's New in Machine Learning

  In this webinar, my colleagues and I discuss the latest announcements from AWS in Machine Learning and Artificial Intelligence.

• Let the Computer Enforce It For You

  As a programming newbie, you’re often told to comment on your code as much as possible. But it’s not long before you read an article telling you to do otherwise. Confused? This post will give you a clear picture of when to comment and when to avoid it.

• SaaS Tenant Isolation in AWS Machine Learning

  Machine Learning Software-as-a-Service providers with a lot of tenants need to keep these tenants’ prediction endpoints separate. Here are several ways to do that, balancing expense and robustness. We show a way to do flexible access control to the Sagemaker inferencing endpoints with an advanced use of Identity and Access management.

  See the article “Implementing SaaS Tenant Isolation Using Amazon SageMaker Endpoints and IAM”

  Also, see the talk “Multitenant inference architecture with SageMaker endpoints”.

• Cloud Blaster

  When experimenting with Google Cloud Platform, your playground environment tends to get cluttered. This post describes a tool to fix that.

• Why I work at DoiT as a Cloud Infrastructure Consultant

  I have been working at DoiT International as a Senior Cloud Architect since 2019. Maybe you should too?

  More at my article here.

• From Notebook to AWS

  You typically start learning machine learning with Notebooks, but for production use you will need to move to distributed services. This post walks through the AWS equivalents for Notebook functionality.

• AWS Artificial Intelligence/Machine Learning Black Belt

  I completed the AWS Artificial Intelligence/Machine Learning Black Belt, an advanced certification with more than a multiple choice example: It requires a final project roughly at the level of a university term project. See the blog post here.

• Kotlin, Gradle, and the Cloud

  I have wanted to use Kotlin for a long time. Though it is better known for use with Android, Kotlin is a general purpose language: Java as it would be if it could gain from 20 years of lessons learned and leave behind old decisions. This blog post gives the fast track to getting started.

• Resource Labeling with Iris3

  Iris 3 adds labels to resources like VM instances, Pub Sub Topics, and more, making for more detailed cost reports. See the blog post here.

• The Quickest Quickstarts

  When I want to use a new technology, I first want a script that gives me a working “Hello World”; I can tweak from there. This article presents nine such solutions for different compute infrastructures on AWS and GCP.

• AWS Firewalls: How and When to Use Each One

  When I saw that AWS had a new firewall with the uninformative name “Network Firewall,” I thought “not another one.” This is my attempt to sort them all out. Then Jeff Barr, AWS Chief Evangelist, tweeted it!

• Acausal Trade

  I am proud of this article that I had a big part in writing, which was now chosen by LessWrong as a “gem” worthy of reposting.

  I have not contributed anything new to the concept of acausal trade, but it is the first article I know about the topic: Previously there were nothing but hints in LessWrong comments. This article has been cited a few times. There it is possibly it is still the best article on the topic.

  Thanks to Michael Vassar who explained the topic to me in 2012. That understanding inspired me to write the article

• Looking for an emulator for Google Cloud Tasks?

  Though an emulator is often requested, Google does not offer one for developing with Cloud Tasks, comparable to what if offers for Datastore or PubSub. I created one.

  More about it at my blog post, or go straight to GitHub.

• Slate Star Codex Online Meetups

  Please see here for the list of past meetups (with video); and future meetups.

• The hidden costs of Datastore

  Datastore export costs don’t show up in Google Cloud Monitoring. Here’s how to set up real-time alerts to better keep track of export costs.

  More on how to detect this at my article.

• Safe Scrub

  Safely delete the resources cluttering your Dev and QA projects.

  Details, with open-source code, at my article.

• Too Many Threads in Cloud Pub/Sub

  On first use, the Google PubSub Java client creates 60 threads that stay live permanently.

• You can handle pods, but what about clusters?

  Kubernetes makes it easy to orchestrate the pods that run your applications. But what about the clusters that the pods run on? These need a lot more hands-on care to configure correctly across the clouds, and in particular, to clone a configuration of a cluster from one cloud to another.

  In this article, I compare cluster models across Azure, Google, and Amazon and present a tool for cloning clusters within one cloud or cross-cloud.

  More at my article here.

• Copying Data in BigQuery

  If you want to copy data between BigQuery tables, there are many ways to do it depending on whether you want

  • Free
  • Inter-regional
  • Automated
  • Partitioned tables

  More at my article here.

• How Your Web App Can Serve the Chinese Market

  For web app developers, serving users in China requires a completely different way of thinking. I’ll explain how to work in practical steps to achieve good-enough solutions.

  More at my article here.

• How to Best Prepare for your Cloud Certification Exam

  All about the best preparation materials for Google Cloud Architect and AWS Cloud Architect Associate and Professional exams.

  More at my article here.

• Build on Your Experience to Earn Cloud Certifications

  The AWS and Google Cloud Platform certifications exams are tough, but your years of experience will help you. Not through your knowledge, because no one knows everything in the exam curriculum, but through your other professional strengths.

  More at my article here.

• Google shuffles your SSH keys

  You try to SSH to a VM in Google Compute Engine, and get “WARNING: REMOTE HOST IDENTIFICATION HAS CHANGED!”

• Cultish, I wish

  What is corporate loyalty?

• Society for the Protection of Obnoxious People