Oldrich Vykydal

API Gateway + Kubernetes 🤝. Kong Inc. Champion and Chief Integration Evangelist / Chief Architect at OPITZ CONSULTING Deutschland GmbH, Sven Bernhardt on how to secure APIs with plugins in your Kubernetes management.  Full blog + video here covering:  - 🛠️ Installing Kong Ingress Controller - 🔐 Securing APIs with plugins - 📡 Exposing Kubernetes services as APIs, securing APIs with plugins (authentication, rate-limiting) - 🌐 Using the Gateway API for HTTP/TCP routing - 📊Observing analytics - all managed via the centralized Kong Konnect control plane. #KongPin #Kubernetes #Ingress #K8s #LearnToCode

1

A well put together incident report. I wonder if they used chatgpt to write it. "The Kubernetes data plane can operate largely independently of the control plane, but DNS relies on the control plane – services don’t know how to contact one another without the Kubernetes control plane. In short, the root cause was a new telemetry service configuration that unexpectedly generated massive Kubernetes API load across large clusters, overwhelming the control plane and breaking DNS-based service discovery."

2

2 Comments

Yeah, even I am guilty of building my foundations on top of the UNIX philosophy, programming in C (the one without pluses) and even C++ with its highly optimized object-oriented data model. Just a few days ago, I learned about yet another petty Linux / UNIX design trap that can surprise you. It's the output stream buffering in the standard C library on UNIX systems that also affects C++. In the standard library, from the UNIX design point of view an integral part of the operating system, the decision on default buffering is made based on what the output of a program is connected to. For interactive programs it is important that you see the output lines as they are emitted by the application code. Therefore the default is line buffering where a newline character flushes the buffer and the data is passed to the terminal – but only when the output file descriptor is known to be a terminal. Programs that don't talk to a terminal do not flush every line by default and thus do not produce all output if they crash before flushing the buffer. I learned about this while using Compiler Explorer when I realized that it's not possible to debug crashing programs with simple `std::println` (from the C++23 <print> header) or similar tools (writing to `std::cout`) because the output is never shown. Of course there are solutions to the problem including those as simple as using `std::println` together with `std::cerr` or changing the buffering configuration of `std::cout`. However, my first suspect was the Compiler Explorer itself, not an age old “feature” of the libc. And this can easily bite you even as a non-developer if you pipe programs output to any interactive viewer with the assumption that the program has a text output that is line-buffered.

8

12 Comments

Do you really need to glamorize using Kubernetes? What if your application can be managed with Docker itself? Let's delve into why and when we might choose Kubernetes over Docker. #Docker #Kubernetes https://lnkd.in/ge7Fm7sP

2

AzAPI: Cutting-edge access to Azure APIs It provides a thinner layer, allowing for direct access to the latest Azure API versions as soon as they're available.

2

𝗛𝗮𝗽𝗽𝘆 𝟭𝟬𝘁𝗵 𝗕𝗶𝗿𝘁𝗵𝗱𝗮𝘆 𝗞𝘂𝗯𝗲𝗿𝗻𝗲𝘁𝗲𝘀! 🎉   It’s been 10 years since an open source project from a scrappy team at Google burst onto the scene and became a global phenomenon. A decade later, Kubernetes has become the go-to way to run large applications in the cloud.   𝗪𝗵𝗮𝘁 𝗶𝘀 𝗞𝘂𝗯𝗲𝗿𝗻𝗲𝘁𝗲𝘀?   It's a container orchestration platform that handles services like load balancing, releases and rollback, and scaling, among other infra management tasks. It allows management of containerised applications as a collection of microservices and represents an alternative to virtual machines, the other common way of packaging computing environments.   K8s has become the second largest open source project and has dominated the market, surpassed only by Linux!   If you want to see how far Kubernetes has come over the past 10 years, James Spurin from DiveInto shared an interactive, hands-on, in-browser version of the very first version of Kubernetes (v1.0.0). The lab is accessible through his Github repository (🔗link in comments) and you can run it absolutely for 𝗙𝗥𝗘𝗘! Kubernetes has come a long way since the first ever official release of the project, and this is a cool way to celebrate the evolution of the project.   HBD K8s 🎂 #kubernetes #k8s #containerization #devops

49

1 Comment

💕 ...I like Go now, I guess I "finished" my latest project `Go Distributed Cache`, which is a highly scalable, fault-tolerant distributed caching system written in Go, designed to handle high availability, strong consistency, and scalability using quorum-based replication and sharded caches. This project was a deep dive into Go, and I was truly surprised by how concise and productive this language is. The built-in benchmarking and profiling tools made it an absolute joy to optimize and test the system. 🎯Key Features: - Sharded Caches for better performance and reduced contention. - gRPC Communication for efficient and fast inter-node messaging. - Quorum-Based Replication to ensure strong consistency, even during node failures. - Dynamic Membership that allows nodes to join and leave the cluster seamlessly. - Consistent Hashing for efficient key distribution and auto-balancing across nodes. Though it’s a non-production system at this point, the project explores key concepts in distributed systems while leveraging the power of Go. 👉 Check it out here: https://lnkd.in/espEKVaJ #GoLang #DistributedSystems #OpenSource

13

Why you are most likely using Helm wrong when using it in a #GitOps setup What I see a lot of companies who use #Helm do, is to define the version (or tag) of the #docker image in the actual helm chart. Some hardcode it in the #template file, others in the default values.yaml file. And then when they need to release their software component, they create a new release of their Helm chart before releasing their software component. This mean that you have to make a change in two repositories. You'll need a new release of your Helm chart and one of your service. This is wrong in so many ways. If you are using #Flux you produce a #HelmRelease file. This file containers a reference to whatever value file in the chart you want to use as your default values, and you can add your own values that overload the default values. Your helm chart should only have a new release if there are changes to the actual templating part. In this section you can define the image tag / version. Like this :

31

4 Comments

Regarding databases, it is not a secret that I love #postgres and #mongodb. It goes without saying that I'm giddy about #postgres.new, an in-browser postgres sandbox with a chatbot, backed by #pglite.dev (postgres packaged as #wasm). I can think of many use cases for pglite, especially in terms of moving compute directly to the client's browser. checkout 👇 short intro video https://lnkd.in/e2VcyWTY

18

2 Comments

Keeping your #kubernetes clusters lean is easier than you may think. Join Matthew Parker and me tomorrow to learn how.

2

What a great post about the realities of widely adopted OSS projects. Read all the comments. As a contributor to several OSS projects with wide reach, it’s surprising to see Helm in such a place. I can’t speak highly enough of the documentation (and yet even it draws some criticism). I absolutely love to see the community manager engaging in the discussion as well. The importance of contribution and architecture documentation is highlighted. As one project I contribute to had its reach exponentially grow, those coming to the project had expectations the project would solve their problems, had advice that failed to meet the opinionated goals, wanted to steer the project in bespoke directions, or opened issues that really lacked a first look at the documentation. And never mind failing to follow through after a tough review of an MR. My own effort was delivered after a thorough review and my own tenacious belief of the value to the community. I’m left with a “how to help” itch, but I’m stretched as it is and unfamiliar with Go. Helm is such a great tool. I can only offer moral support and raise awareness. Maybe the CNCF can steal a chapter or two from the ASF runbook.

7

🚀 **Exploring FastAPI: Building a Backend with Python's Type Hints** FastAPI has become one of my go-to tools for building APIs, thanks to its speed, simplicity, and support for modern Python features like type hints. 🌟 In my latest backend project, I leveraged FastAPI to handle everything from data validation to auto-generating interactive API documentation. **Why FastAPI?** - **Speed**: FastAPI is one of the fastest frameworks out there, perfect for performance-critical applications. - **Type Hints**: Python’s type hints offer better code completion and ensure data validation out of the box. - **Minimal Code**: You get more functionality with fewer lines of code, making development faster and cleaner. - **Interactive API Docs**: FastAPI automatically generates Swagger UI docs, which makes testing endpoints easier during development. Here’s a quick snapshot of what I built: - **CRUD API**: Implemented create, read, update, and delete functionalities with type-safe operations. - **Data Models**: Used Python's `pydantic` library for defining models and ensuring validation at runtime. - Payment system for Paystack, Flutterwave, PayPal and stripe It’s incredible to see how much time FastAPI saves, especially when handling complex API structures. I highly recommend it to anyone working on API-driven backend projects and you can always and you can always dm to develop your backend architecture using Fast API with python. I’ll share more about my experience and insights on this soon! 💡 #FastAPI #Python #BackendDevelopment #TypeHints #APIDevelopment #WebDevelopment #TechInnovations

5

"Kafka is dead." Strong words, but I get the sentiment. Kafka's been the undisputed king of streaming for years. But it's showing its age. Not cloud native, data agnostic, and a pain to manage at scale. The Kafka API, though, that's the real gem. It's become the lingua franca of data streaming, and it's fueling a wave of innovation around Kafka, not within it. Confluent’s Kora is a step in the right direction, but the real excitement is in the ecosystem: Redpanda Data, WarpStream, Estuary Flow... these are the tools pushing the boundaries of what's possible with streaming data. So, is Kafka dead? Not yet. But its reign is definitely challenged. The future belongs to those who embrace cloud native, data centric approaches. And that might not be Kafka itself, but the vibrant ecosystem it inspired.

1

How much can MVP infrastructure on AWS cost? $200, +-$50. This is the approximate number for a small MVP like Amor's Secret which we recently pushed to production (https://lnkd.in/drMzXe2B). Considering you have a small mobile app, a non-rocket-science-based backend a few services integrated here and there, that's the number for AWS hosting. Having it higher is a question to dig into.

4

🏗️🚀 The Backstage Scaffolder is an essential tool for driving Golden Paths across engineering teams. We’ve been working on documenting, for the first time, all available Scaffolder actions and the parameters that feed information into them. Our goal? To make life easier for the Backstage community and simplify template creation from here on out. Take a look at what’s possible with the Scaffolder on the link below 👇 #backstage #scaffolder #goldenpath

1

Made an OpenApi (re)generator (api first) for Scala. It generates source files and then *refactors your existing code*, not overwriting it blindly in the target folder.. Directly spitting out code you would have written by hand! 🙂 Currently it only supports my sharaf framework. I have in plan to add http4s and others along the way. Try and let me know what you think! Source: https://lnkd.in/egixKjnT Demo: https://lnkd.in/ezXYf3g4 #scala #openapi #swagger

18

I speed up the processing of my Flink Application operations by increasing the Runtime Environment Parallelism. The FlatMap operator tasks parallelism is set to executionEnvironment default parallelism. You can base on the execution Environment parallelism in this context kubernetes you can configure the parallelism for your source, processing topology and the sink operator tasks. You can dynamically get the runtime execution environment value in java with the StreamExecutionEnvironment class. Flink on Kubernetes is a Game changer. Similar to how GitLab runner on kubernetes change the game of dynamic CI/CD runners. #cloudengineering #cloudcomputng #softwareengineering #dataengineering #distributedsystem #kubernetes #coding #programming

14

Chatgpt is becoming more of a great everyday tool in an IT person's toolbox. The more you understand it and understand how to extract the information you need out of it, its easy to skip pass the fluff or things that dont make sense. Needing to import hundreds of existing subnets into terraform using terraforms import block, i knew i did not want to import all of that per subnet and that i also needed to import these into modules. Essentially I have a code block that does a for_each through all of the subnets that come into the variable. import {  for_each = { for key, value in var.subnet : key => value if value.import }  id = each.value.import_resource_id  to = module.subnet.azurerm_subnet.main[each.key] } Only problem was i needed all of those subnets in a specific format in my tfvars file before i could do that. In comes chatgpt and a very much needed python script. As somebody who didnt have the need or time to spend a ton of time writing that script out, i was able to ask it a very small subset of questions piecing this script together found here in my github https://lnkd.in/gWSarEpH. I of course used powershell to export out the subnets in a specific format also that was needed. All in i was able to take a task that could have taken me days to complete by not having to manually import all those subnets in one by one and also generating a script using chatgpt vs having to search forums and ask around and get a bunch of methods for the same outcome. What are some IT use cases you have used chatgpt to improve daily task? #chatgpt #ai #automation #scripting

5

1 Comment