Coordinated Restore at Checkpoint [En]

⚠️ Cette présentation se tiendra en anglais car l'intervenant n'est pas francophone.

Imagine your Java application starting up in just 5 milliseconds. Sounds incredible, right? Almost too good to be true? And no, this isn't about native images.

There's an innovative solution called Coordinated Restore at Checkpoint (CRaC).
Initially, this concept emerged from the Linux CRIU project: Checkpoint/Restore In Userspace.
However, CRIU doesn't always deliver reliable results. That's where CRaC comes in, offering a JVM-specific implementation.

In this talk, we'll look into the complexities of CRIU and after that, we will turn to CRaC.
We will discuss how CRaC operates within the Java ecosystem, its integration with Spring, and how you can use it in your own applications for near-instant startup times.
From this talk you will learn what is possible to CRaC and what is not, what are the benefits and limitations, and what is the cost of CRaC-ing.

---
Pasha Finkelshteyn is a Developer Advocate for BellSoft, he educates the public on the latest software tools built by the company and helps to create and test instruments for data engineers.
When not advocating, he writes in Kotlin, speaks at conferences, composes a new article, or maintains the Kotlin API for Apache Spark.