In this blog article I’ll show you how you can use the confluent CLI to set up a Kafka cluster on Confluent Cloud, the necessary API keys, and then a managed connector. The connector I’m setting up is the HTTP Source (v2) connector.

It’s part of a pipeline that I’m working on to pull in a feed of data from the UK Environment Agency for processing. The data is spread across three endpoints, and one of the nice features of the HTTP Source (v2) connector is that one connector can pull data from more than one endpoint.

Much as I love kcat (🤫 it’ll always be kafkacat to me…), this morning I nearly fell out with it 👇

😖 I thought I was going stir crazy, after listing topics on a broker and seeing topics from a different broker.

😵 WTF 😵

ksqlDB is a fantastically powerful tool for processing and analysing streams of data in Apache Kafka. But sometimes, you just want a quick way to profile the data in a topic in Kafka. I wrote about this previously with a convoluted (but effective) set of bash commands pipelined together to perform a GROUP BY on data. Then someone introduced me to visidata, which makes it all a lot quicker!

Whilst Apache Kafka is an event streaming platform designed for, well, streams of events, it’s perfectly valid to use it as a store of data which perhaps changes only occasionally (or even never). I’m thinking here of reference data (lookup data) that’s used to enrich regular streams of events.

You might well get your reference data from a database where it resides and do so effectively using CDC - but sometimes it comes down to those pesky CSV files that we all know and love/hate. Simple, awful, but effective. I wrote previously about loading CSV data into Kafka from files that are updated frequently, but here I want to look at CSV files that are not changing. Kafka Connect simplifies getting data in to (and out of) Kafka but even Kafka Connect becomes a bit of an overhead when you just have a single file that you want to load into a topic and then never deal with again. I spent this afternoon wrangling with a couple of CSV-ish files, and building on my previous article about neat tricks you can do in bash with data, I have some more to share with you here :)

One of the fun things about working with data over the years is learning how to use the tools of the day—but also learning to fall back on the tools that are always there for you - and one of those is bash and its wonderful library of shell tools.

I’ve been playing around with a new data source recently, and needed to understand more about its structure. Within a single stream there were multiple message types.

👉 Ingesting XML data into Kafka - Introduction

What would a blog post on rmoff.net be if it didn’t include the dirty hack option? 😁

The secret to dirty hacks is that they are often rather effective and when needs must, they can suffice. If you’re prototyping and need to JFDI, a dirty hack is just fine. If you’re looking for code to run in Production, then a dirty hack probably is not fine.

One of my favourite hacks for getting data into Kafka is using kafkacat and stdin, often from jq. You can see this in action with Wi-Fi data, IoT data, and data from a REST endpoint. This is fine for getting values into a Kafka message - but Kafka messages are key/value, and being able to specify a key is can often be important.

Here’s a way to do that, using a separator and some jq magic. Note that at the moment kafkacat only supports single byte separator characters, so you need to choose carefully. If you pick a separator that also appears in your data, it’s possibly going to have unintended consequences.

There’s ways, and then there’s ways, to count the number of records/events/messages in a Kafka topic. Most of them are potentially inaccurate, or inefficient, or both. Here’s one that falls into the potentially inefficient category, using kafkacat to read all the messages and pipe to wc which with the -l will tell you how many lines there are, and since each message is a line, how many messages you have in the Kafka topic:

I’ve been playing around with the new SerDes (serialisers/deserialisers) that shipped with Confluent Platform 5.5 - Protobuf, and JSON Schema (these were added to the existing support for Avro). The serialisers (and associated Kafka Connect converters) take a payload and serialise it into bytes for sending to Kafka, and I was interested in what those bytes look like. For that I used my favourite Kafka swiss-army knife: kafkacat.

kafkacat is one of my go-to tools when working with Kafka. It’s a producer and consumer, but also a swiss-army knife of debugging and troubleshooting capabilities. So when I built a new Fedora server recently, I needed to get it installed. Unfortunately there’s no pre-packed install available on yum, so here’s how to do it manually.

I’ve been poking around recently with capturing Wi-Fi packet data and streaming it into Apache Kafka, from where I’m processing and analysing it. Kafka itself is rock-solid - because I’m using ☁️Confluent Cloud and someone else worries about provisioning it, scaling it, and keeping it running for me. But whilst Kafka works just great, my side of the setup—tshark running on a Raspberry Pi—is less than stable. For whatever reason it sometimes stalls and I have to restart the Raspberry Pi and restart the capture process.

Wi-fi is now ubiquitous in most populated areas, and the way the devices communicate leaves a lot of 'digital exhaust'. Usually a computer will have a Wi-Fi device that’s configured to connect to a given network, but often these devices can be configured instead to pick up the background Wi-Fi chatter of surrounding devices.

There are good reasons—and bad—for doing this. Just like taking apart equipment to understand how it works teaches us things, so being able to dissect and examine protocol traffic lets us learn about this. However, by collecting this type of traffic it can be possible to track and analyse behaviour in ways that we may or may not feel comfortable with. Improving public transport? Sure. Tracking shopper behaviour? Meh, less sure.

ksqlDB 0.7 will add support for message keys as primitive data types beyond just STRING (which is all we’ve had to date). That means that Kafka messages are going to be much easier to work with, and require less wrangling to get into the form in which you need them. Take an example of a database table that you’ve ingested into a Kafka topic, and want to join to a stream of events. Previously you’d have had to take the Kafka topic into which the table had been ingested and run a ksqlDB processor to re-key the messages such that ksqlDB could join on them. Friends, I am here to tell you that this is no longer needed!

You can download the InfluxDB connector for Kafka Connect here. Documentation for it is here.

When a message from your source Kafka topic is written to InfluxDB the InfluxDB values are set thus:

I’m quite a fan of Sonos audio equipment but recently had some trouble with some of the devices glitching and even cutting out whilst playing. Under the covers Sonos stuff is running Linux (of course) and exposes some diagnostics through a rudimentary frontend that you can access at http://<sonos player IP>:1400/support/review:

Whilst this gives you the current state, you can’t get historical data on it. It felt like the problems were happening "all the time", but were they actually? For that, we need some cold, hard, data! Something like this, in fact:

This is based on using Confluent Cloud to provide your managed Kafka and Schema Registry. All that you run yourself is the Kafka Connect worker.

Optionally, you can use this Docker Compose to run the worker and a sample MySQL database.

Kafka Connect in distributed mode uses Kafka itself to persist the offsets of any source connectors. This is a great way to do things as it means that you can easily add more workers, rebuild existing ones, etc without having to worry about where the state is persisted. I personally always recommend using distributed mode, even if just for a single worker instance - it just makes things easier, and more standard. Watch my talk online here to understand more about this. If you want to reset the offset of a source connector then you can do so by very carefully modifying the data in the Kafka topic itself.

Kafka Connect has as REST API through which all config should be done, including removing connectors that have been created. Sometimes though, you might have reason to want to manually do this—and since Kafka Connect running in distributed mode uses Kafka as its persistent data store, you can achieve this by manually writing to the topic yourself.