Florent Ramière

@bbejeck @kafkastreams @apachekafka Congrats @bbejeck, quite an awesome achievement!

I'm excited to share that Kafka Streams in Action 2nd edition has arrived! Hopefully, it will be a useful resource for people interested in developing stream-processing applications. manning.com/books/kafka-st… @kafkastreams @apachekafka

Excited to launch @getconduktor 2.0 on @ProductHunt today! We've revamped the UI for a more intuitive experience and expanded the free tier with powerful features. Take a look and share your feedback! producthunt.com/posts/conduktor

Le premier meetup 2024 de l'OSDI (Open Source Data Infrastructure) aura lieu le 6 mars chez @Google Paris 🤩, merci de nous accueillir ! On vous a concocté un programme aux petits oignons ! Faites vite ! Les places sont limitées. On aura @framiere et @Tundjii comme speakers ! 📝 meetup.com/paris-open-sou…

@IngoKegel you seem to have plenty of 404 on your Jprofiler website.

Is your Kafka infrastructure eating into your budget? Try Conduktor Gateway to trim down costs!

Our latest blog is about clarifying your Kafka costs before they burn you! 🌶️❤️‍🔥 It's time for a cost analysis journey! Spotlight on: multi-tenancy, caching, compression, best practices, common pitfalls, safeguarding, encryption, compliance, large records. conduktor.io/blog/8-new-way…

@avinash_ukr @getconduktor It is a different capability, Gateway is all about the native Kafka protocol. So you can enhance your existing applications without any change!

@getconduktor Is this a rest proxy alternative or a different capability?

🚀 Thrilled to unveil Conduktor Gateway at #KafkaSummit2023 to address common #ApacheKafka challenges: ✅ Enforcing best practices ✅ Streamlined Kafka testing ✅ Field-level encryption ✅ RBAC & auditing Discover our journey to create Conduktor Gateway: bit.ly/41tCvC8

@gcuisinier conduktor.io obviously! Check out all the proxy features too: they're awesome!

Any alternative to Lenses to visualise / manage / debug kafka cluster?

@vtimd @getconduktor Welcome to the team @vtimd!

Starting the new year fresh as the Principal Developer Advocate for @getconduktor! Let’s see what kind of trouble we can get into!

@C0urante @rmoff @Mr_mitchellh Perfect, I remember giggling while writing it ;)

@framiere @rmoff @Mr_mitchellh I just threw up a little

Tell me you've been writing #shell scripts without telling me you've been writing shell scripts…

@Mr_mitchellh @rmoff You can even fake kafka in bash ;) github.com/framiere/kafka…

@rmoff The best language ever created.

@AlcaideSantos @riferrei @apachekafka In my experience, they are very often misunderstood or worst abused.

We all ❤️ @apachekafka, but what are Kafka's weakest points for you? What are your biggest pains? Give me your best shots!

@jaiminks @apachekafka I am not sure what you are referring to. Kafka transactions span multiple topics. What do you need to do manually, or what is slowing you down?

@framiere @apachekafka Kafka does not support out of box transactions support for multiple tables as a part of single unit of work.

@grahamstirling @apachekafka @bufbuild While we're on SR, the magic byte + 4 bytes used by the AvroSerializer is a massive pain. I understand it has been created prior to the introduction of headers. But we should have a smarter implementation that does the things everybody expects.

@grahamstirling @apachekafka @bufbuild Oh, ok. I fully agree, SR is a bit clunky/rusty/oldy. Confluent is investing in very interesting governance features based on SR. It's pretty interesting, check it out. Also, Conduktor/AKHQ can alleviate the SR/UI pain a bit.

@cloudnativeguru @bibryam @apachekafka Performance at scale, which scale? Technical, or organizational? Partitions is indeed a pain. Solace has plenty of great features. NATs is a specific beast to me. And for perf/latencies, RedPanda has interesting properties where they shine.

@framiere @bibryam @apachekafka Based on Java. Bad performance at scale. Explicit partitions. Go NATS or Solace.

@davgarcia5 @gunnarmorling @apachekafka 2015: "I like this idea a lot and the change is quite simple and natural." @jaykreps 2022: Status Open. This kip solves a ton of things EOS can't. 😥

@framiere @gunnarmorling @apachekafka This: issues.apache.org/jira/browse/KA…

@odedia @dunithd @apachekafka @gwenshap By 'client-side complexity', do you mean its configuration, its performance consideration, idempotency understanding, commit strategies, replay, error handling, rebalance listeners, ... ? Tell me more :)

@framiere @dunithd @apachekafka I think the client-side complexity. I remember a tweet from @gwenshap that said (at a high level, correct me if I’m wrong, Gwen) that perhaps a simple Rest-type of client would have made more sense in retrospect. It exposes critical performance-altering params to developers.

@GKlijs @apachekafka Agreed, let me rephrase my question then: What are Kafka's weakest points when Kafka is the right tool for the job? :)

@framiere @apachekafka I think Kafka is fine, it's no Golden hammer, and that's a good thing. It's bad at some things, and that's fine.