by gunnarmorling
0 subcomment
- To those who flagged this submission, can you share your reasoning? What guidelines do you see being violated by it? Thanks!
Disclaimer: I'm the author of this post.
by ekjhgkejhgk
2 subcomments
- "Considered harmful" considered harmful.
by cultofmetatron
1 subcomments
- Id argue that if you are in the position where you legitimately NEED kafka, you hopefully also know what you're doing. You're outside the audience for the "just use postres" crowd. That said, if you're in a startup with a few thousand users, just use postgres is still solid advice.
by wewewedxfgdf
0 subcomment
- There's many many ways to make a message queue these days - all the main SQL databases can act as a queue - everything from Postgres to MS SQL server to MySQL to Oracle to sqlite to the custom applications like Kafka and for the most part they are all more or less valid - it's not all about Postgres.
Take the approach that appeals to you and feel happy about it without big open source telling you "you're holding it wrong!"
- In my experience, "you don't need *MQ, just use Kafka" is a way worse problem.
Trying to explain the distinction between an event streaming platform and a distributed message queue to your enterprise architect is an exercise that no one should have to go through.
by brettgriffin
1 subcomments
- > Looking to make it to the front page of HackerNews?
Nailed it. I read the original post earlier this week and was very impressed with its technical detail. But the point of the the post was incongruent with the post's title. But the post got way more attention because of that title.
But if you think about the effort it took to write that post, the title was a really good bet on ROI.
by scottcodie
1 subcomments
- One thing the other blog post missed and this post misses too is that you don't need Kafka to use Debezium with Postgres. This gives you a pretty seamless onramp to event streaming tools as you scale.
by philipwhiuk
1 subcomments
- > Named a Java Champion, I enjoy speaking at conferences, for instance at QCon, JavaOne, Red Hat Summit, JavaZone, JavaLand and Kafka Summit.
by oompydoompy74
0 subcomment
- Insufferable tone aside, I really dislike the “right tool for the job” argument. The correct tool is the one that is handy and gets the job done. Has the author never encountered a Swiss Army Knife?
by jauntywundrkind
0 subcomment
- I'm more interested in the "You don't need Kafka the product, when we have this Kafka protocol compatible alternative". Kafka is more than a product: it's become a standard, with many many implementations. I'd love to see wider coverage of the alternatives. RedPanda, StreamNative Ursa, OSO, Aiven, many others.
by hactually
2 subcomments
- isn't Kafka old news at this point?
LinkedIn have moved onto Northguard... but no GitHub yet
by blindriver
2 subcomments
- ""You don't need Kafka" considered harmful by employees of Kafka."
- Could we please just agree not to use this "considered harmful" phrase to describe advice where the answer is "depends"? This kinda makes the author seem like he has lost the ability to consider what software is out there. That he is working for Kafka doesn't help.
Example: Someone writes a software that could use something simple like SQLite, and they switched to Postgres for performance reasons. Now unless what Kafka beings is the core reason they switched to Postgres not pulling in another dependency and adding a nother piece to the puzzle, can be a total legitimate engineering decision. And that renders the "considered harmful" utterly ridiculous.
Use a system like Kafka if you need what it brings (a distributed event streaming platform). If that isn't what you need or a very simple postgres solition suffices, go for that. Maybe you need event streaming but distributing it is overkill. Maybe you just need some sort of queue. Who knows? Not the author of this post.