jvm-serializers

Compared with protobuf, fury is 3.2x faster. When comparing with avro, fury is 5.3x faster. Compared with flatbuffers, fury is 4.8x faster. See https://github.com/eishay/jvm-serializers/wiki for detailed benchmark data

There are too many java serialization libraries, I compared fury with jdk/kryo/fst/protostuff/protobuf/flatbuffers/thrift/msgpack/canproto/acro/jackson/json. Fury are fastest too. See https://github.com/eishay/jvm-serializers/wiki for detailed benchmark results. It's a hard choice to select one for the title, so I use JDK for it, which may be not a good choice. BTW, fury support jit serialization for jdk17 record, which is super fast compared to other serialization frameworks such as kryo

Fury is also the fastest jvm serialization framework in the https://github.com/eishay/jvm-serializers/wiki

> I'm surprised this is up for debate. I looked into logging in protobuf when I was seeing if there was a better binary encoding for ring-buffer logging, along the same lines as nanolog: https://tersesystems.com/blog/2020/11/26/queryable-logging-w... What I found was that it's typically not the binary encoding vs string encoding that makes a difference. The biggest factors are "is there a predefined schema", "is there a precompiler that will generate code for this schema", and "what is the complexity of the output format". With that in mind, if you are dealing with chaotic semi-structured data, JSON is pretty good, and actually faster than some binary encodings: https://github.com/eishay/jvm-serializers/wiki/Newer-Results...

do you have any links to benchmarks showing json serialization inefficiency for large sets of numbers? I found this which shows json isn't too much worse than others, but it doesn't go into extreme cases https://github.com/eishay/jvm-serializers/wiki

Gson best of the breed? Oh my :( https://github.com/eishay/jvm-serializers/wiki https://github.com/fabienrenaud/java-json-benchmark https://github.com/kostya/benchmarks

Some interesting benchmarks of various Java serialization libraries: https://github.com/eishay/jvm-serializers/wiki

HTTP+JSON can actually be very quick if implemented correctly. And most importantly it is easy to debug which is critical when you are starting off. JSON can be faster than binary in many cases: https://github.com/eishay/jvm-serializers And your sentence sounds like the ideal scenario, no ?

For JVM most popular benchmark is https://github.com/eishay/jvm-serializers/wiki

Easy there tiger ;) What does it even mean that something is string based? Here: https://github.com/eishay/jvm-serializers/wiki a JVM bench where JSON is pretty competitive with binary formats [and results would be event better if project owners actually merged PRs ;( ]

Check out the performance of these serializers: https://github.com/eishay/jvm-serializers/wiki Although they're mostly limited to Java, I know at least one (Avro) can be used in Python.

The two are not really comparable since you get different results (JSONObjects are not strongly typed). You will easily be able to find workloads where one is better than the other, but in the average case, ig-json-parser will probably be faster because it's based on Jackson. (see json/jackson/manual vs json/org.json/manual on https://github.com/eishay/jvm-serializers/wiki)

Hi! I'm one of the authors of this library. The code this library emits is very similar to the json/jackson/manual serializer/deserializer in https://github.com/eishay/jvm-serializers/wiki. Conveniently, GSON is benchmarked there too. You're right, the field names could be guessed at compile time. Please feel free to submit a PR! Thanks!

Or "I'm making an educated guess". Otherwise how can you get truly surprised when things turn out to be different in the real world :-) I'd also guess it would be more expensive during decoding. So let's compare that with some real data: https://github.com/eishay/jvm-serializers/wiki (old wiki but with some graphs https://code.google.com/p/thrift-protobuf-compare/wiki/Bench...) it turns out that state of the art JSON serialisers aren't so bad after all, especially the serialisation.