Athar Al Najjar’s Post

Calling all data pros interested in Kafka and event-driven architectures: Get ready to embark on a journey of learning and discovery, all wrapped in the guise of a comic book. Ada and Jax’s escapades provide the perfect backdrop, but the true gem is the opportunity to learn the fundamentals of Apache Kafka and Confluent. It’s not just a story (although the story is pretty good 😉 ); it’s a tool to unlock your understanding of real-time data processing. Check out Confluent’s first ever comic book “The Data Streaming Revolution: Rise of the Kafka Heroes”-- afterall who doesn’t love fun and learning? Get your copy now!

Calling all data pros interested in Kafka and event-driven architectures: Get ready to embark on a journey of learning and discovery, all wrapped in the guise of a comic book. Ada and Jax’s escapades provide the perfect backdrop, but the true gem is the opportunity to learn the fundamentals of Apache Kafka and Confluent. It’s not just a story (although the story is pretty good 😉 ); it’s a tool to unlock your understanding of real-time data processing. Check out Confluent’s first ever comic book “The Data Streaming Revolution: Rise of the Kafka Heroes”-- afterall who doesn’t love fun and learning? Get your copy now!

It’s baaack. Confluent’s second issue of Confluent Chronicles is here! Inside, you’ll find a clever, but relatable approach (if I do say so myself) to learn about the fundamentals of Apache Flink, stream processing, and Confluent. Our heroes, developer and architect team, Ada and Jax are struggling with decentralized data processing challenges and need a manageable, cost-effective stream processing solution for an important upcoming launch. They embark on their next adventure to learn why Apache Kafka and Apache Flink are better together. Get your copy now!

It’s baaack. Confluent’s second issue of Confluent Chronicles is here! Inside, you’ll find a clever, but relatable approach (if I do say so myself) to learn about the fundamentals of Apache Flink, stream processing, and Confluent. Our heroes, developer and architect team, Ada and Jax are struggling with decentralized data processing challenges and need a manageable, cost-effective stream processing solution for an important upcoming launch. They embark on their next adventure to learn why Apache Kafka and Apache Flink are better together. Get your copy now!

It’s baaack. Confluent’s second issue of Confluent Chronicles is here! Inside, you’ll find a clever, but relatable approach (if I do say so myself) to learn about the fundamentals of Apache Flink, stream processing, and Confluent. Our heroes, developer and architect team, Ada and Jax are struggling with decentralized data processing challenges and need a manageable, cost-effective stream processing solution for an important upcoming launch. They embark on their next adventure to learn why Apache Kafka and Apache Flink are better together. Get your copy now!

It’s baaack. Confluent’s second issue of Confluent Chronicles is here! Inside, you’ll find a clever, but relatable approach (if I do say so myself) to learn about the fundamentals of Apache Flink, stream processing, and Confluent. Our heroes, developer and architect team, Ada and Jax are struggling with decentralized data processing challenges and need a manageable, cost-effective stream processing solution for an important upcoming launch. They embark on their next adventure to learn why Apache Kafka and Apache Flink are better together. Get your copy now!

It’s baaack. Confluent’s second issue of Confluent Chronicles is here! Inside, you’ll find a clever, but relatable approach (if I do say so myself) to learn about the fundamentals of Apache Flink, stream processing, and Confluent. Our heroes, developer and architect team, Ada and Jax are struggling with decentralized data processing challenges and need a manageable, cost-effective stream processing solution for an important upcoming launch. They embark on their next adventure to learn why Apache Kafka and Apache Flink are better together. Get your copy now!

It’s baaack. Confluent’s second issue of Confluent Chronicles is here! Inside, you’ll find a clever, but relatable approach (if I do say so myself) to learn about the fundamentals of Apache Flink, stream processing, and Confluent. Our heroes, developer and architect team, Ada and Jax are struggling with decentralized data processing challenges and need a manageable, cost-effective stream processing solution for an important upcoming launch. They embark on their next adventure to learn why Apache Kafka and Apache Flink are better together. Get your copy now!

Check out Confluent’s second issue of Chronicles! Inside, you’ll find a clever, but relatable approach to learn about the fundamentals of Apache Flink, stream processing, and Confluent. Our heroes, developer and architect team, Ada and Jax are struggling with decentralized data processing challenges and need a manageable, cost-effective stream processing solution for an important upcoming launch. They embark on their next adventure to learn why Apache Kafka and Apache Flink are better together. Get your copy now!

This is a fantastic writeup. I believe that application code should be agnostic of where it runs (i.e. not have a hard dependency on Kubernetes, yarn, etc). Much of Samza's complexity, as you mention, stems from the fact that it required a deployment orchestrator (not workflow orchestrator like LittleHorse😉 🐴) such as YARN or Mesos to run. #kafkastreams is "just a library" (don't tell Matthias I said that), which makes it far more powerful since it can be used anywhere without opening a ticket for the infrastructure team. The fact that it's only dependency is—to quote Sophie Blee-Goldman—"a JVM and an #apachekafka cluster" is a great strength. I, too, am skeptical about the wisdom of Flink's choice to embed code that talks to YARN and also the K8s API directly into Flink. Why? It is just more dependencies to keep track of, more API's to maintain, etc...which will cost resources (which are preciously scarce in open-source software dev). More things can break during release testing, more release blockers, slower releases...etc. In addition, I think that "ability to get running quickly" is a huge differentiator for open-source projects. I once had the pleasure of speaking with an early sales engineer at Mongo. He said "the number one metric we tracked was how long it took for someone to download Mongo and run their first query." In LittleHorse, you can start with a completely fresh laptop, get LH running with just a single docker container, and run your first workflow in under 8 minutes. Try it here in #python: https://lnkd.in/gCyZssqU