Loïc Mathieu
loicmathieu.fr
Loïc Mathieu
@loicmathieu.fr
Lead Developer @kestra.io, Java & JVM passionate, performance, OSS contributor (mainly Quarkus), Book Author (Quarkus edition ENI), GCP Google Dev Expert.
Like to learn stuff and share what I learned.
https://www.loicmathieu.fr
Reposted by Loïc Mathieu
Cette semaine, j'ai testé l'outil 'mise-en-place' (MEEZ ahn plahs avec l'accent 🇬🇧), et je suis déjà conquis 🤯
Je vais pouvoir remplacer tous mes scripts custom moches que j'avais du mal à maintenir 😅

Mon avis et ce que j'ai bidouillé est à lire ici : codeka.io/2025/12/19/a...
Adieu `direnv`, Bonjour `mise` - Julien Wittouck
freelance solution & software architect 🏗 - containers 🐋 & linux 🐧 ❤️ - teacher & trainer 🎓 @ univ-lille.fr - speaker 🎙 - Team @Cloud_Nord
codeka.io
December 19, 2025 at 5:14 PM
Reposted by Loïc Mathieu
Kubernetes 1.35 have been released 🎉

With 3 releases each year, it cannot be easy to know the new features, changes and enhancements.

Let's understand Kubernetes 1.35, Timbernetes, with this changelog in a visual way:
dev.to/aurelievache...

@cncf.io
Understanding Kubernetes: part 60 – Kubernetes 1.35 Changelog
Serie of sketchnotes about Kubernetes. Explaining in a visual way Kubernetes principles.
dev.to
December 19, 2025 at 11:35 AM
Reposted by Loïc Mathieu
Docker Hardened Images (DHI) is now free and open source (Apache 2.0) 🎉

- hardened, minimal, near-zero CVEs images & helm charts
- SBOM & provenance attestations

www.docker.com/blog/docker-...

@docker.com #SupplyChainSecurity
Hardened Images for Everyone | Docker
Security for everyone. Docker Hardened Images are now free to use, share, and build on with no licensing surprises.
www.docker.com
December 17, 2025 at 2:14 PM
Reposted by Loïc Mathieu
In this week's blog post, you'll learn that JFR doesn't like arrays in other complex types as fields in custom JFR events, and what types are supported: mostlynerdless.de/blog/2025/12...
Don't use Arrays or other Complex Types in Custom JFR Events - Mostly nerdless
Learn about arrays and complex types in custom JFR events and about all the supported field types in this weeks blog post.
mostlynerdless.de
December 12, 2025 at 7:49 AM
Not having transactions didn't mean you couldn't implement a lock!

This is how we implement a lock mechanism in Elasticsearch inside Kestra.
www.loicmathieu.fr/wordpress/in...
Implementing a lock in Elasticsearch | Loic's Blog
At Kestra, the workflow orchestration and scheduling platform I work for, we have two repository implementations: JDBC (with support for H2, MySQL, and Postgres) and Elasticsearch. In JDBC, locking a…
www.loicmathieu.fr
December 11, 2025 at 11:41 AM
Reposted by Loïc Mathieu
#Google announcing official #MCP support for various Google and #GoogleCloud services
(#BigQuery, #GKE, security, #GoogleMaps...)
cloud.google.com/blog/product...
December 11, 2025 at 9:40 AM
Reposted by Loïc Mathieu
Why write engineering blogs? Here’s how some of our favorite bloggers responded to the question “Why did you start blogging and why do you continue?” writethatblog.substack.com/p/why-write-...
December 10, 2025 at 3:29 PM
Reposted by Loïc Mathieu
Interesting times at @azulsystems.bsky.social
December 11, 2025 at 10:29 AM
Reposted by Loïc Mathieu
Expand your #Java applications' performance with this new guide on optimizing #MongoDB queries by Farhan Hasin Chowdhury. Boost efficiency and enhance load times now! #Java #MongoDB #Optimization foojay.io/today/optimi...
Optimizing MongoDB Queries in Java Applications
Modern Java applications often struggle with performance bottlenecks that have little to do with the JVM itself. In most cases, the culprit lies deeper in how the application interacts with its databa...
foojay.io
December 10, 2025 at 6:44 AM
Reposted by Loïc Mathieu
Finding the right memory size for Java apps is tricky but vital for performance. 🧩 @frankdelporte.be’s blog shows how GC logs, JMeter tests, and Azul vs OpenJDK builds reveal key insights. Optimize cost + speed here: https://bit.ly/4pQ0S9H

#Java #Performance
Learn More About the Memory Use of Java Applications
Determining the ideal memory size for Java applications can be challenging, but critical to achieve maximum performance.
bit.ly
December 6, 2025 at 4:45 PM
Reposted by Loïc Mathieu
📝 Blogged: "You Gotta Push If You Wanna Pull"

Wrote up some thoughts on push vs. pull queries, materialized views, and why data duplication isn't something to fear.

👉 www.morling.dev/blog/you-got...
December 7, 2025 at 4:28 PM
Reposted by Loïc Mathieu
Pretty impressive vector indexing improvements in Elasticsearch by leveraging GPUs! (via @benwtrent.bsky.social) www.elastic.co/search-labs/...
Elasticsearch GPU: vector indexing acceleration with NVIDIA - Elasticsearch Labs
Discover how Elasticsearch achieves nearly 12x higher indexing throughput with GPU-accelerated vector indexing and NVIDIA cuVS.
www.elastic.co
December 5, 2025 at 9:13 AM
Reposted by Loïc Mathieu
JDK 26 has been branched from the main line and is now in Rampdown Phase One: mail.openjdk.org/pipermail/jd...

openjdk.org/projects/jdk/26/

#JDK26 #OpenJDK #Java
mail.openjdk.org
December 4, 2025 at 6:53 PM
Reposted by Loïc Mathieu
"Mini-LSM--Build a simple key-value storage engine in a week"

This looks like a great learning resource for folks looking to get started with log-structured merge trees in #Rust 🦀, by Alex Chi Z.

👉 skyzh.github.io/mini-lsm/
December 4, 2025 at 2:14 PM
Reposted by Loïc Mathieu
Et si vous pouviez contribuer au programme Devoxx 2026 ? 👀

Rejoignez les ✨ Community Reviewers ✨ pour évaluer les propositions soumises, et aider à identifier les tendances émergentes !

Pour participer, inscrivez-vous via ce formulaire ⭢ docs.google.com/forms/d/e/1F...

À très vite 👋
December 4, 2025 at 9:09 AM
Reposted by Loïc Mathieu
Discover the exciting features of Java 25 in this comprehensive overview by @loicmathieu.fr! Set the pace and adapt to change more easily with #Java25. Check it out here: foojay.io/today/java-2...
🚀 #Foojay #JavaUpdate #DeveloperTools
Java 25: What's New?
foojay is the place for all OpenJDK Update Release Information. Learn More.
foojay.io
December 3, 2025 at 1:30 PM
Reposted by Loïc Mathieu
Bonjour, pour rappel, l'ouverture des partenariats se fait aujourd'hui à 14h ! Rendez vous sur notre site partenaire.devlille.fr pour faire votre demande !

Si vous souhaitez plus d'informations, n'hésitez pas à aller voir notre site ou à nous envoyer un mail à contact@devlille.fr !
December 1, 2025 at 9:33 AM
Reposted by Loïc Mathieu
🤔 Have you ever wondered how #kubernetes caches #etcd?

✍️ I just published a new deep-dive into K8s watch cache: per-resource Cachers, waitUntilFreshAndBlock(), and progress notifications for quiet resources.

pierrezemb.fr/posts/diving...
Diving into Kubernetes' Watch Cache
Understanding how Kubernetes apiserver caches etcd, the 3-second timeout, and K8s 1.34 consistent read feature
pierrezemb.fr
November 12, 2025 at 8:13 AM
Reposted by Loïc Mathieu
ELT or ETL?

ETL transforms before loading, so what lands is already clean and compliant. ELT loads raw first, then transforms in the warehouse, which is faster and more flexible at cloud scale.

Kestra orchestrates either way: youtu.be/VxOmetLBTxE
ETL vs ELT: What's The Difference?
YouTube video by Kestra
youtu.be
November 27, 2025 at 4:23 PM
Reposted by Loïc Mathieu
New Token-Oriented Object Notation (#TOON) Hopes to Cut #LLM #Costs by Reducing Token Consumption

https://www.infoq.com/ne...
November 27, 2025 at 5:30 PM
Reposted by Loïc Mathieu
Explore the Art of Performance Tuning! Learn how saving in the #cloud is pointless if your code isn't optimized, from Daniel Witkowski, who highlights that an efficient cloud means nothing if you're wasting code 1000× more.

#CodeEffeciency #PerformanceTuning

foojay.io/today/the-ar...
The Art of Performance Tuning: Why Saving 30% in the Cloud Means Nothing if Your Code Wastes 1000× More
foojay is the place for all OpenJDK Update Release Information. Learn More.
foojay.io
November 28, 2025 at 9:45 AM
TIL that you can create indices concurrently in Postgres using `CREATE INDEX CONCURRENTLY ` and in MySQL with `ALGORITHM=INPLACE, LOCK=NONE`.
This is tremendously convenient when you use automatic migration on multi-hundred GB tables that happen at startup and don't want this startup to take hours!
November 28, 2025 at 8:51 AM
Reposted by Loïc Mathieu
📝 Blogged: "On Idempotency Keys"

Discussing several options for ensuring exactly-once processing in distributed systems using idempotency keys, from UUIDs to monotonically increasing sequences.

👉 www.morling.dev/blog/on-idem...
November 25, 2025 at 4:38 PM