HAProxy Technologies

HAProxy Fusion 2.0 is LIVE! This release marks a generational leap, delivering a unified security control plane and hyperscale automation. We've introduced improved security management, a re-engineered API for high-performance automation, and native orchestration of HAProxy Enterprise’s new Threat Detection Engine. Read the PR: hubs.la/Q046WxbH0

Certificate validity periods are getting shorter and proposals are already in motion to reduce them to as little as 47 days. Manual renewal workflows are not going to scale with that pace. The ACME protocol automates the full certificate lifecycle, from domain validation to issuance and revocation, with no custom scripts or manual intervention. And HAProxy supports it natively from version 3.2 onward. Read more: hubs.la/Q04gP2bT0

At 23 million requests per second, CPU efficiency is a crucial capacity decision. @criteo 's SRE, Basha Mougamadou, knows this better than most. HAProxy's auto-computed CPU policy reduced context switching by 20%, freeing up meaningful CPU headroom for actual request traffic. Read more: haproxy.com/blog/how-crite…

The ACME protocol was built to take operational burdens, such as managing hundreds or thousands of TLS certificates manually, off your plate entirely. By automating certificate issuance, renewal, and revocation at the proxy layer, teams get a single, consistent approach to certificate management across every application. HAProxy 3.2 and HAProxy Enterprise 3.2 introduced native ACME support, with two challenge types available: HTTP-01 for domain verification via a well-known URL path, and DNS-01 for teams that need wildcard certificates or cannot expose HTTP endpoints. From HAProxy 3.3 onward, certificate saving is fully automated through event notifications and the HAProxy Data Plane API — no manual Runtime API calls required. Read the full configuration walkthrough: hubs.la/Q04gNcMp0

The HAProxy Enterprise Kubernetes Ingress Controller handles tens of thousands of RPS on a single instance — highest throughput, lowest latency, lowest CPU usage, and an embedded WAF that secures your clusters without added complexity. Deploy inside or outside of Kubernetes, scale across multiple clusters, and manage traffic the way your team already works. Find out more: hubs.la/Q04gCPLf0

If you feel like modern infrastructure changes so fast your load balancer can't keep up, this live session is for you. Join Ron Northcutt, Director of Technical Marketing, on May 28th for a practical, hands-on session on the HAProxy Data Plane API. You'll walk away with: 🔹A working mental model of how the Data Plane API thinks about HAProxy 🔹Live demos covering server registration, runtime map updates, and zero-downtime SSL certificate rotation 🔹Scripting that handles version conflicts the way production automation should 🔹A clear framework for when to reach for the API and when to reach for a control plane 📆 May 28 ⏲️11AM ET / 5PM CET Register here: hubs.la/Q04gB3fP0

Ever since @WillyTarreau wrote the first HAProxy code in 2001, the established philosophy and principles of HAProxy have never changed: performance, reliability, and flexibility must be balanced. After 25+ years without a rip-and-replace moment, HAProxy's gradual, community-driven evolution from standalone proxy to fully programmable application delivery platform is a blueprint worth understanding. Read the full article: haproxy.com/blog/how-hapro…

This is exactly the kind of brilliant innovation that reminds me why HAProxy exists. The SRE team at @criteo didn't just scale their infrastructure. They reimagined how a load balancer should work at scale, building a fully automated LBaaS platform on HAProxy that handles everything dynamically while their engineers can focus on building. Inspiring story, check it out: haproxy.com/blog/how-crite… @HAProxy

Most #Kubernetes Gateway API controllers ask you to choose between standards compliance, performance, and operational simplicity. HAProxy Unified Gateway 1.0 was built to remove that trade-off entirely. This free and open source release is built on HAProxy 3.2 LTS — the same core that routes traffic for some of the most demanding infrastructure on the internet. It covers Gateway API specifications 1.3, 1.4, and 1.5, supports HTTP, HTTPS, and TLS routing, and introduces three new CRDs that bring full HAProxy configuration management natively into Kubernetes. Dynamic scaling no longer requires a reload. Pod changes are reconciled directly via the Runtime API, keeping active connections alive throughout every scale event. Prometheus metrics and session persistence ship out of the box. Helm chart support makes installation and upgrades a standard, repeatable operation. With Ingress NGINX permanently archived, the migration question is no longer optional. HAProxy Unified Gateway gives platform teams a replacement that doesn't just fill the gap but raises the bar entirely. Check out the full breakdown: hubs.la/Q04ghL5Q0

Nehmen Sie an unserer Live-Sprechstunde teil (nur live, ohne Aufzeichnung) und erfahren Sie, wie führende Unternehmen ihre Kosten um bis zu 90 % senken. Wir sprechen nicht nur über Einsparungen – wir zeigen Ihnen konkret, wie eine Migration technisch umgesetzt wird: mit maximaler Skalierbarkeit und moderner Security. Jetzt anmelden: hubs.la/Q04gcR290

If you're building on #AWS and working through challenges around scale, resilience, or application security, join us at the #AWSSummit in NYC! We'll be at booth 915 where the team will present HAProxy One in action, practical load balancing strategies, cloud-native application delivery, performance optimization, and security within AWS environments. Whether you're early in your cloud architecture journey or deep into a complex deployment, we'd love to hear what you're working on. You can set up a meeting with the team here: hubs.la/Q04g7bwn0 See you in New York!

Rejoignez notre session de questions-réponses en direct (sans enregistrement) pour découvrir comment les entreprises leaders réduisent leur coût total de possession (TCO) jusqu’à 90 %. Rejoignez-nous : haproxy.com/fr/content-lib…

When your baseline is 23 million requests per second across six data centers, manual configuration is not an option. See how Criteo used the HAProxy Runtime API to build a fully automated LBaaS platform on HAProxy: hubs.la/Q04fYxCC0

Latency is one of those terms everyone uses and fewer people define precisely. It's not just "slowness." Latency is the time between a request being made and a response being received. In distributed systems, it compounds across every hop, every service, and every network boundary between your user and your application. Understanding where latency originates is the difference between chasing the wrong bottleneck for weeks and fixing the right one in hours. Read the full breakdown: hubs.la/Q04fYmFt0

$5.2 million for the next F5 upgrade was the moment DoubleVerify decided enough was enough. By switching to HAProxy Enterprise, they now process over 5 billion requests per day at a 90% TCO cost reduction — with full observability, horizontal scalability, and zero downtime upgrades. Find out how they got there: hubs.la/Q04fPZyn0

The best teams I've worked with recognized their breaking point before the business did. @criteo is one of them. See how they built a fully automated LBaaS platform on HAProxy, enabling them to achieve 23 million requests per second and automate renewal for thousands of TLS certificates all without a single manual intervention. Full story: haproxy.com/blog/how-crite…

Last chance to register for today's F5 migration office hours. ⌛ Join us live and find out how organizations are cutting infrastructure TCO by up to 90% and how your team can do the same. Register now: hubs.la/Q04fMB-P0

Gaming platforms don't get much bigger than Roblox. Over 111.8 million daily active users, millions of requests per second, hundreds of load balancer instances across the globe. But this also means unpredictable traffic spikes that come with hosting 6.4 million user-created experiences. 📈 When COVID-19 sent their user numbers through the roof, their F5 hardware load balancers couldn't keep up. They needed something built for that kind of scale and that kind of pressure. They utilized HAProxy Enterprise to enable self-service routing, zero-reload dynamic updates, and a WAF integration that brought their security team peace of mind. Read the full story: hubs.la/Q04fCH2Q0

If your team is managing deployments across multiple clouds, this session is for you. Watch this on-demand session and see how Universal Mesh empowers you to simplify, scale, and secure modern applications in any environment. Find out how to connect everything, solve "unsolvable" IP conflicts, eliminate the "mesh tax", centralize observability, and more. Watch here: hubs.la/Q04frg8p0

HAProxy Enterprise 3.3 is out now, marking a meaningful step forward for teams running HAProxy Enterprise WAF and bot management at scale. - HAProxy Enterprise 3.3 introduces a unified WAF module that consolidates both the Intelligent WAF Engine and CRS compatibility mode under a single configuration interface. - ALOHA 18.0 is also out, bringing the same CRS v4 migration flexibility to virtual and hardware appliance deployments. Read more: haproxy.com/blog/announcin…