DevOps Industry Updates #25

25 issues just like that! What started as an update segment at my Meetup has turned into something bigger. Thanks to all my loyal readers out there, I’ve heard you loud and clear: you love the rich technical content and lack of vendor spam. True to that brand, issue #25 is loaded with only the most impactful stories and has a theme of debugging systems at scale, including gPRC, JVM and PostgreSQL. Ready to open a bunch of new tabs? Here we go:

🔥 Top Cream

This issue’s top 4 stories:

  1. Why load balancing gRPC is tricky?
  2. Tricks of the Trade: Tuning JVM Memory for Large-scale Services
  3. Debugging random slow writes in PostgreSQL
  4. The Full Story of the Stunning RSA Hack Can Finally Be Told

🌎 Society

  • Vamp is joining CircleCI!: “together we will combine continuous integration, continuous deployment, release orchestration and continuous validation, and bring it to the next level. Our unified platform will enable software developers, engineers, DevOps teams, and business stakeholders alike to “shift-right”, and deliver better software even faster and more frequently.”

  • The Next Step after DevOps and GitOps Is Cloud Engineering, Pulumi Says: if we are going to treat infrastructure as code, shouldn’t infrastructure engineers have access to the same tools that make software engineers productive and even the same languages? That’s the theory behind Pulumi, which has just released version 3 of its open source platform.

  • My thoughts about the Principal role

📟 DevOps

  • Tricks of the Trade: Tuning JVM Memory for Large-scale Services: Uber’s growth over the last few years exponentially increased both the volume of data and the associated access loads required to process it, resulting in much more memory consumption from services. Increased memory consumption exposed a variety of issues, including long garbage collection (GC) pauses, memory corruption, out-of-memory (OOM) exceptions, and memory leaks.

🛠️ DevOps Tools

  • Sublime Text 4: new features include a redesigned UI, Apple Silicon support, tab multi-select, context-aware auto complete, GPU rendering and much more!

  • terraform-docs: generate documentation from Terraform modules in various output formats.

☸️ Kubernetes

  • Kubernetes capacity planning: How to rightsize your cluster: don’t be greedy! Learn how to identify unused resources and how to rightsize the capacity of your Kubernetes clusters.

  • Service Mesh Wars, Goodbye Istio: after using Istio in production for almost 2 years, we’re saying goodbye to it. Learn why, as well as the current state of the Service Mesh Wars.

  • Introducing PodTopologySpread: managing Pods distribution across a cluster is hard. There is a common need to distribute Pods evenly across topologies, so as to achieve better cluster utilization and high availability of applications. The PodTopologySpread scheduling plugin (originally proposed as EvenPodsSpread) was designed to fill that need.

  • Database migrations on Kubernetes using Helm hooks: leverage the pre-install and pre-upgrade Helm hooks to run database migrations before your application is installed or updated. Keep reading to understand why simpler solutions might not be the best idea and a couple of gotchas when using Helm hooks.

  • Using Finalizers to Control Deletion: deleting objects in Kubernetes can be challenging. You may think you’ve deleted something, only to find it still persists. While issuing a kubectl delete command and hoping for the best might work for day-to-day operations, understanding how Kubernetes delete commands operate will help you understand why some objects linger after deletion.

🔐 Security

💻 Programming

📖 Machine Learning

  • MUM: A new AI milestone for understanding information: Google’s Multitask Unified Model is trained across 75 different languages and can understand many different tasks at once, allowing it to develop a more comprehensive understanding of information and world knowledge than previous models. MUM is multimodal, so it understands information across text and images and, in the future, can expand to more modalities like video and audio.

🐧 Linux

🚢 Leadership

☁️ Cloud

  • Why load balancing gRPC is tricky?: using a binary protocol with structured data as the communication medium among services is indeed attractive, but there are some considerations when using gRPC, most important of all is how to handle load balancing.

  • Service-Oriented vs. Monolith: most teams do choose the microservices path since that’s the “industry standard” these days. However, monolithic designs still have their use and space, especially at an early stage of an idea or a product.

  • The Architecture of Uber’s API gateway: Uber developed a feature-rich API gateway that is capable of complex operations on the incoming and outgoing data payload across multiple protocols. This article takes a deeper dive into the technical components of Uber’s custom API gateway system.



Article version: 1.0.1

Written on May 23, 2021