Scale to Zero with Spring + GraalVM or WebAssembly by Sébastien Deleuze
Ғылым және технология
Recording brought to you by American Express. americanexpress.io/kotlin-jobs
Cloud can be expensive, complex and consume a lot of resources due to server instances waiting for incoming requests. To avoid those issues, Sébastien is going to share 2 ways of deploying your Kotlin web applications in a scale-to-zero fashion with more flexibility than Serverless platforms like Amazon Lambda.
Sébastien will talk about the GraalVM native support introduced by Spring Boot 3.0 which allows you to compile your Kotlin/JVM applications into a native executables that start instantly, have a reduced memory consumption and can produce optimized container images.
But Sébastien is also exploring how WebAssembly could be used instead of traditional container images as a way to deploy Kotlin workloads to the Cloud and the Edge in a scale-to-zero fashion. This is possible with Kotlin/Wasm, the incubating Kotlin support for WebAssembly initially designed to run in the browser.
You will see it in action combined with KoWasm, an experimental project created by Sébastien, in order to build server-side applications leveraging WASI (Web Assembly System Interface).
Talk by: Sébastien Deleuze
#Kotlinconf23 #spring #Kotlin #webassembly #KotlinConf
Пікірлер: 2
Have been waiting for this talk of Sebastian. Great!
Great overview of all the progress in the kotlin wasm area. I greatly appreciate all the hard work Sebastian has been doing to advocate progress on this front. As a backend developer that does kotlin, kotlin multi-platform, and supports a kotlin-js UI, I'm very much looking forward to a more unified stack and WASM is the big unifying enabler in that space. From my point of view, the multi-platform library ecosystem is actually rapidly maturing. There are a few missing things here and there. But overall, I'm able to do lots of interesting things with it and the amount of jvm dependencies that I use, or miss, is rapidly decreasing. A big blocker with current wasm support is that most multi-platform libraries are not yet compiling to wasm ... which means that particularly there is not a whole lot you can actually do with Kotlin wasm until that is resolved. For example, my kt-search library should be something that could work on WASM as well (and it would make a lot of sense in e.g. serverless/edge networking). But it will require things like co-routines, ktor-client, datetime, and a few other things to be available. I hope to see lots of progress on that front this year and I think getting some early support for this should be a big goal for Kotlin 2.x and set up the beginnings of the future Kotlin ecosystem that Sebastian sketches out in this presentation. Once that is in place, a lot of stuff could be ported over relatively quickly. For my own library, I've started tracking 3rdparty issues for wasm support as blockers: github.com/jillesvangurp/kt-search/issues/69. Once those are resolved, I can support wasm pretty quickly.