Apache Flink

hey folks, is this <https://www.amazon.co.uk/Stream-Processing-Apache-Fabian-Hueske/dp/149197429X/ref=sr_1_3?crid=G9MK8TZZTB9N&amp;keywords=flink&amp;qid=1692959745&amp;sprefix=flink%2Caps%2C123&amp;sr=8-3|book on flink> still up to date/a relevant read?

thought it might be outdated given how much i've seen change over the last year or so

The Flink docs are actually pretty good, but I’v just recently realized how complicated Flink is.

If someone wrote a scientific paper on how checkpoints work in stream processing, I’d read it. :smile: 

put this one together :smile:
<https://leanpub.com/streamprocessingwithapacheflink>

<@U04BH4L796Y> also putting together stuff on it.
taking you from barrier alignment to unaligned checkpoints, buffer debloating, incremental checkpoints and the changelog state backend

:thinking_face: will have a look! is this more recent?

it was published during the summer :slightly_smiling_face:

and will get more updates in a few months, which you will get for free after

I’ll ask my boss to pay for that. It should be exactly what we need right now.

Is there something somewhere that explains more on how to actually measure the queue between two subtasks? I’d like to know how many and how big.

Another thing I want to know is where the Java UDF spends its time, and how often is its ‘state’ stored and retrieved, and if it’s java serialized when.

Stream Processing with Apache Flink is still relevant, especially the first few chapters if you’re new to stream processing.

It mostly covers the DataStream API after which hasn’t changed much.

<@U03LXQ4FUQZ> Your book already increased the performance of our aggregation jobs several hundred percent! :sweat_smile: 

really excited to hear this.. and im working on more advanced and in-depth stuff :smile: