Nicolas Fränkel

My ideas for blog posts inevitably start to dry up after over two years at Apache APISIX. Hence, I did some triage on the APISIX repo. I stumbled upon this one question: We have a requirement to use a plugin, where we need to route the traffic on percentage basis. I’ll give an example for better understanding. We have an URL https://xyz.com/ca/fr/index.html where ca is country (canada) and fr is french language. Now the traffic needs to routed 10% to https://xyz.com/ca/en/index.html a

I continue to work on my Opentelemetry demo. Its main idea is to showcase traces across various technology stacks, including asynchronous communication via an MQTT queue. This week, I added a couple of components and changed the architecture. Here are some noteworthy learnings; note that some of them might not be entirely connected to OpenTelemetry. Here’s an updated diagram. New components appear in violet, and updated components appear in green. I want to be able to add more comp

I’m in the process of adding more components to my OpenTelemetry demo (again!). The new design deploys several warehouse services behind the inventory service so the latter can query the former for data via their respective HTTP interface. I implemented each warehouse on top of a different technology stack. This way, I can show OpenTelemetry traces across several stacks. Anyone should be able to add a warehouse in their favorite tech stack if it returns the correct JSON payload to the inv

A (long) time ago, my first job consisted of implementing workflows using the Staffware engine. In short, a workflow comprises tasks; an automated task delegates to code, while a manual task requires somebody to do something and mark it as done. Then, it proceeds to the next task - or tasks. Here’s a sample workflow: The above diagram uses Business Process Model and Notation. You can now design your workflow using BPMN and run it with compatible workflow engines. Time has passed. Sta

I’ve always liked GUI, both desktop-based and browser-based before you needed five years of training on the latter. That’s the reason I loved, and still love Vaadin: you can develop web UIs without writing a single line of HTML, JavaScript, and CSS. I’m still interested in the subject; a couple of years ago, I analyzed the state of JVM desktop frameworks. I also like the Rust programming language a lot. Tauri is a Rust-based framework for building desktop applications. Here&

I try to constantly deepen my knowledge of HTTP and REST. Recently, I stumbled upon the list of all registered HTTP Headers. This post is dedicated to the Vary HTTP Header. The problem Two years ago, I wrote about web resource caching server-side. The idea is to set up a component between the client and the upstream to cache previously computed results to avoid overloading the latter. Depending on your infrastructure and requirements, this component can be a reverse proxy or an API Gateway. H

I recently read 6 Ways To Pass Parameters to Spring REST API. Though the title is a bit misleading, as it’s unrelated to REST, it does an excellent job listing all ways to send parameters to a Spring application. I want to do the same for Apache APISIX; it’s beneficial when you write a custom plugin. General setup The general setup uses Docker Compose and static configuration. I’ll have one plugin per way to pass parameters. docker-compose.yml services: httpbin: ima

I wrote previously about libs for error management in Rust. This week, I want to write about the try block, an experimental feature. The limit of the ? operator Please check the above article for a complete refresher on error management in general and the ? operator in particular. In short, ? allows to hook into a function call that returns a Result: If the Result contains a value, it continues normallyIf it contains an error, it short-circuits and returns the Result to the calling function.

Last week, I wrote an analysis of the IETF Idempotency-Key specification. The specification aims to avoid duplicated requests. In short, the idea is for the client to send a unique key along with the request: If the server doesn’t know the key, it proceeds as usual and then stores the responseIf the server knows the key, it short-circuits any further processing and immediately returns the stored response This post shows how to implement it with Apache APISIX. Overview Before starting

The first rule of distributed systems is 'Don’t distribute your system'. Designing distributed systems right is infamously hard for multiple reasons. The idempotency concept For example, a call to a function can succeed or fail in non-distributed systems. Once you move the called function to a remote component, a third option appears: you call the remote function but get no response from the component. At this point, it’s impossible to know whether the call reached the component o