2025-11-04 10:00
Building images for cloud and on-premises servers poses challenges like reducing attack surfaces and ensuring compliance. 🌐 Red Hat Lightspeed simplifies image hardening by using OpenSCAP remediations at build time, allowing for pre-hardened images. Users can build images via an API or a user interface on sandbox.redhat.com. 🛠️ The article details steps for selecting the PCI-DSS profile and customizing images, ensuring compliance throughout the process. For compliance monitoring, Red Hat...
Source: Red Hat Developer Blog
Gianluca Zuccarelli, Obinna Ezeakachi
2025-11-04 07:01
Post-training methods are crucial for enhancing large language models (LLMs) beyond their initial pre-training phase. 🌐 These methods include supervised fine-tuning, continual learning, and reinforcement learning, each aimed at making models more useful and aligned with specific tasks. Techniques like Orthogonal Subspace Fine-Tuning and Parameter-Efficient Fine-Tuning optimize memory and performance. 📈 Developers can explore these approaches using the open-source Training Hub library, which...
Source: Red Hat Developer Blog
Mustafa Eyceoz, Addie Stevens
2025-11-03 08:01
Unlock the power of eBPF in Linux! 🚀 This article explores how eBPF can help attribute packet drops to specific netfilter rules in the Linux kernel. It dives into the netfilter subsystem, illustrating how to pinpoint which firewall rule caused a packet drop and how to hook into kernel processing for detailed insights. Learn how to create a simple nftables ruleset and utilize eBPF tools for effective troubleshooting. 🔍💻 #eBPF #LinuxKernel #Netfilter #Firewall #PacketDrops
Source: Red Hat Developer Blog
Toke Høiland-Jørgensen
2025-11-03 08:01
Minimizing bandwidth during system updates is crucial. The article discusses how the **rpm-ostree build-chunked-oci tool** helps by isolating related rpm packages into separate layers. This means that changes to one configuration file won't require downloading large amounts of data. A new feature allows specific files to be assigned to particular layers, improving efficiency, especially for large embedded images. Setting up a test environment and using this tool can significantly speed up...
Source: Red Hat Developer Blog
Chris Kyrouac
2025-11-03 07:01
🚀 Deploying large language models (LLMs) on Red Hat OpenShift AI enhances on-premise inference for the Ansible Lightspeed intelligent assistant. You can containerize, scale, and integrate LLM workloads. This ensures better data control and compliance with organizational policies. Key components include model storage, a serving runtime, and GPU infrastructure. Performance is evaluated using Time to First Token (TTFT) and Inter-Token Latency (ITL) to maintain a responsive user experience. 🔗...
Source: Red Hat Developer Blog
Riya Sharma, Elijah DeLee
2025-10-30 13:02
Organizations transitioning to AI production face critical decisions on inference platforms. vLLM, a library of open-source code, optimizes large language model (LLM) performance through efficient GPU memory use. Its architecture, including advanced KV-Cache management and parallelization strategies, supports diverse hardware and enhances scalability. As an open-source project under the PyTorch Foundation, vLLM ensures sustainable innovation and flexibility, making it a strong choice for...
Source: Red Hat Developer Blog
Fatih E. Nar, Greg Pereira, Yuan Tang, Robert Shaw, Anish Asthana
2025-10-30 12:46
🎉 Happy 1st Birthday, Repo! 🎉 This year marked the debut of Repo, our helpful "reponaut," who assists developers in managing complex coding challenges. Key events included advancements in Red Hat OpenShift AI and the successful growth of the Quarkus project, highlighting the focus on developer productivity. In 2025, major releases such as Red Hat Enterprise Linux 10 and the Red Hat OpenShift Lightspeed assistant enhanced the developer experience. The community continues to embrace AI and...
Source: Red Hat Developer Blog
Caroline King, Colleen Lobner
2025-10-30 07:01
🔧 Red Hat OpenShift 4.20 introduces the `oc adm upgrade recommend` command, aimed at simplifying the update process. This tool offers tailored version recommendations to enhance your update planning. 📋 The precheck feature actively identifies alerts that may hinder updates, like the critical `ClusterOperatorDown` alert, ensuring you have the necessary information before proceeding. ✅ This command helps manage risks effectively, allowing for a more confident upgrade experience. Explore the new...
Source: Red Hat Developer Blog
Subin Modeel
2025-10-29 07:01
📊 As AI applications grow, understanding messaging systems like ActiveMQ Artemis and Apache Kafka is essential. ActiveMQ Artemis focuses on transactional reliability and supports various messaging protocols. It's ideal for point-to-point and batch processing. In contrast, Apache Kafka excels in real-time data streaming, offering high throughput and the ability to replay messages for analytics. Choosing between them depends on your specific use case needs. #MessagingSystems #ActiveMQ...
Source: Red Hat Developer Blog
Greg Pack, Prapthi Cariappa Kayapanda
2025-10-27 07:01
Explore the future of AI with RamaLama! 🌐 This open-source CLI simplifies the deployment of vision language models (VLMs) on edge devices. By leveraging container technology, it eliminates complex dependencies, making it easier for developers to implement AI in real-world scenarios. From real-time object recognition to identifying hazards, VLMs are set to revolutionize industries. Learn how to effectively deploy and manage these models with RamaLama. #AI #MachineLearning #EdgeComputing...
Source: Red Hat Developer Blog
Ravishankar Srinivasan
2025-10-27 07:01
🌐 The SDG Hub is transforming how we build synthetic data pipelines for large language models (LLMs). It offers an open framework that enables users to create, compose, and scale data generation with reusable blocks. This modular approach replaces one-off scripts, enhancing reproducibility and innovation. Key features include asynchronous execution, monitoring, and custom block extensions. The SDG Hub supports diverse applications, from knowledge tuning to multilingual data generation....
Source: Red Hat Developer Blog
Aditi Saluja, Abhishek Bhandwaldar, Shivchander Sudalairaj
2025-10-27 07:01
🚀 As enterprises advance AI from experimentation to production, selecting the right hardware accelerators is essential. This article outlines a stage-based framework for choosing AI accelerators throughout the inference lifecycle, covering five key stages: 1️⃣ Initial Setup 2️⃣ Performance Tuning 3️⃣ Production Deployment 4️⃣ Large Model Serving 5️⃣ Edge Deployment Each stage has unique requirements and challenges that influence hardware choices. Understanding these can help organizations...
Source: Red Hat Developer Blog
Christina Zhang
2025-10-24 07:00
Optimize your OpenShift nodes with effective system-reserved parameters! 🚀 This article outlines the steps to calculate and configure CPU and memory resources for underlying components. Proper reservation enhances scheduling and prevents resource overcommitment. Key steps include reserving resources, verifying configurations, and adjusting values for better performance. For more details on implementation, check out the full article! 📊💻 #OpenShift #RedHat #CloudComputing #ContainerManagement...
Source: Red Hat Developer Blog
Sathish Kumar Hemadhri
2025-10-23 07:01
🚨 Important Update for Developers! 🚨 The odo CLI is officially deprecated. The immediate deprecation date is October 23, 2025, with an end of life on March 31, 2026. After this date, no maintenance or support will be provided. For a smoother transition, explore Red Hat OpenShift Dev Spaces for inner-loop development, and consider using Podman and oc CLI for your outer-loop tasks. Stay informed and plan your migration! #RedHat #OpenShift #DevOps #CloudNative #DeveloperTools
Source: Red Hat Developer Blog
Kathryn Yetter
2025-10-23 07:01
🚀 OpenShift users can now enhance their networking capabilities using BGP! This article details how to expose OpenShift networks, providing dynamic routing and advanced features for VMs. Key benefits include automated route configurations, improved performance, and high availability with BFD support. Learn how to export network routes efficiently from OpenShift to a BGP router, ensuring seamless VM migration and connectivity. #OpenShift #BGP #Virtualization #CloudNetworking #FRRouting
Source: Red Hat Developer Blog
Miguel Duarte de Mora Barroso, Valentino Uberti
2025-10-22 07:01
Learn how to effectively run I/O workloads on OpenShift Virtualization VMs! 🚀 This article outlines the steps needed to test I/O performance at scale using a generic FIO workload. It covers prerequisites, such as a functional OpenShift environment and SSH access to virtual machines. Step-by-step instructions detail how to create virtual machines, set up testing parameters, and analyze results, ensuring accurate performance assessments. 📊 Explore the testing process and improve your...
Source: Red Hat Developer Blog
Elvir Kuric
2025-10-22 07:01
Unlock the potential of AI coding with spec-driven development! 🎶 This approach enhances collaboration between humans and AI by establishing clear specifications. Unlike "vibe coding," which can lead to fragile code, spec coding ensures robust results through detailed planning. Stakeholders benefit from early involvement, improving productivity and ROI. By defining the "what" and "how," teams create error-free, scalable code while empowering AI with context. 🤖 Explore how to start with spec...
Source: Red Hat Developer Blog
Rich Naszcyniec
2025-10-21 07:01
Introducing **Krkn-AI**: a new framework for AI-assisted chaos engineering. It addresses the challenges of testing modern systems, especially in dynamic environments like Kubernetes. Chaos engineering helps identify weaknesses by simulating failures, but traditional methods can be manual and static. Krkn-AI automates experiment discovery and execution, allowing teams to focus on insights rather than manual setups. Key features include cluster-aware discoverability, enhanced test coverage, and...
Source: Red Hat Developer Blog
Rahul Shetty, Naga Ravi Chaitanya Elluri
2025-10-21 07:01
Importing provider network routes into Red Hat OpenShift Virtualization via BGP enhances network capabilities. 🌐 This integration allows for dynamic routing and automates configurations, improving VM migration and overall performance. BGP's features like bidirectional forwarding detection (BFD) ensure high availability and rapid failover. To implement this, ensure your OpenShift cluster meets the prerequisites and follow the detailed configuration steps provided. #OpenShift #BGP #CloudNative...
Source: Red Hat Developer Blog
Miguel Duarte de Mora Barroso, Valentino Uberti
2025-10-20 07:01
In the latest analysis of Red Hat OpenShift, a kubelet regression was detected that increased CPU usage by 30% and pod readiness latency by 50%. Our performance engineering team utilized the changepoint detection tool, Orion, to identify these issues during automated scale tests. The regression was linked to kubelet 1.33, which was resolved by reverting to version 1.32.6, restoring normal performance metrics. This case highlights the importance of continuous testing and collaboration to...
Source: Red Hat Developer Blog
Vishnu Challa
2025-10-16 15:44
🚀 Profiling large language models (LLMs) is essential for optimization. This guide details how to set up and profile a vLLM inference server on RHEL with NVIDIA GPUs. 🔧 It covers three main stages: 1️⃣ Environment setup: Install NVIDIA drivers and the Container Toolkit. 2️⃣ Basic profiling: Use the PyTorch profiler to trace inference requests. 3️⃣ Advanced profiling: Leverage NVIDIA Nsight Systems for deeper insights. For a comprehensive understanding, check the full guide! #LLM #GPU #vLLM...
Source: Red Hat Developer Blog
George Zheng Wang
2025-10-16 15:07
🚀 Key findings from a recent study on GPU clusters for distributed training highlight the importance of network architecture. Using IBM Cloud, tests showed that the standard OpenShift pod network creates bottlenecks. For L40S GPUs, secondary vNICs increased performance by up to 132% at scale. For H100 GPUs, switching to SR-IOV led to a 3x increase in throughput. Recommendations emphasize investing in high-performance networks to maximize GPU utilization. #DistributedTraining #GPUPerformance...
Source: Red Hat Developer Blog
Tanya Osokin, Kevin Pouget, Michey Mehta
2025-10-15 07:16
🚀 This October, an update on the Clang bytecode interpreter reveals significant progress! With about 500 commits since last year, the implementation has become more robust. Test failures in the clang suite have decreased from 155 to 90. A working version of `builtin_constant_p` is now available, supporting real-world use cases. Key enhancements include optimizations for constant expressions, improving performance in certain scenarios. The inclusion of libc++ testing has also led to better...
Source: Red Hat Developer Blog
Timm Baeder
2025-10-15 07:16
🚀 Continuous performance testing (CPT) is crucial for software development, especially for Red Hat OpenShift. The article highlights its importance in preventing performance bottlenecks and improving user experience. Challenges include OpenShift's complexity and the need for flexible testing solutions. The team shifted-left, integrating performance tests into their CI/CD pipeline, increasing development velocity and collaboration. Stay tuned for best practices and insights on CPT! 📈🔍 #RedHat...
Source: Red Hat Developer Blog
Joe Talerico
2025-10-14 17:27
Deploying OpenShift in air-gapped environments can be challenging due to complex setup requirements. The article introduces "aba," a new tool that simplifies this process. Aba provides a pre-tested install bundle that includes all necessary components, automating the setup of registries and mirroring images. This tool aims to reduce installation time and manual troubleshooting. Aba's key features include end-to-end automation, built-in best practices, and support for various installation...
Source: Red Hat Developer Blog
Stephen Bylo
2025-10-14 07:01
Unlock the potential of your GPU with NVIDIA's Multi-Instance GPU (MIG) and Red Hat OpenShift! 🚀 This powerful combination allows you to dynamically allocate GPU resources, enabling diverse workloads—from running seven small models to a single large one—without idle time. Discover how the dynamic accelerator slicer operates, facilitating efficient resource management and isolation for teams. 💻 Read more about the setup and see live demos of this innovative technology in action! #NVIDIA...
Source: Red Hat Developer Blog
Harshal Patil
2025-10-13 07:01
Kubernetes provides basic health monitoring for nodes but lacks adequate support for storage and secondary network failures, crucial for virtual machines and telco deployments. The kubelet detects node issues, focusing mainly on resource availability and control plane connectivity. However, it does not monitor storage or network health directly, leading to potential inconsistencies and downtime. To address this, the Node Problem Detector (NPD) operator can be implemented, allowing for...
Source: Red Hat Developer Blog
Raffaele Spazzoli
2025-10-13 07:01
🚀 Exploring GitOps with OCI in OpenShift! Organizations transitioning to GitOps often think they need Git-based systems. However, GitOps principles allow for other compliant storage options like OCI, which can store various content types beyond container images. OpenShift GitOps 1.18 now supports OCI as a source of truth, simplifying management and enhancing security. This approach streamlines operations and integrates well with CI pipelines. To get started, ensure you have the oras CLI and...
Source: Red Hat Developer Blog
Christian Hernandez
2025-10-13 07:00
Unlock the potential of AI with insights-mcp by Red Hat! 🌐 This self-hosted server allows seamless interaction with key Red Hat Insights features like vulnerability management and inventory tracking, all through the Model Context Protocol (MCP). To get started, set up a service account and install VS Code. You can then connect LLM agents to enhance your workflows, from daily security checks to compliance monitoring. Explore the developer preview today and share your feedback! 🛠️💻 #RedHat #AI...
Source: Red Hat Developer Blog
Sean O'Keeffe
2025-10-10 07:00
🚀 Splitting an OpenShift Machine Config Pool (MCP) without node reboots is now possible! This guide details how to create two separate MCPs while maintaining identical configurations. This method is especially beneficial for large clusters with over 100 worker nodes, simplifying upgrades and management. Key steps include identifying the current MCP and machine configurations, creating a new MCP, and labeling nodes for the transition. For detailed commands and procedures, refer to the full...
Source: Red Hat Developer Blog
Rob Fisher
2025-10-10 07:00
Node.js 20 enhances memory management in containers by being container-aware, limiting heap size based on cgroup limits. This adaptation helps prevent memory overflow issues on platforms like OpenShift. The maximum heap size is 50% of the container size, capping at 2 GiB for larger containers. Developers can also set specific limits using the `--max-old-space-size` flag. For efficient CPU allocation, combining `worker_threads` with multiple CPU limits can improve performance, but balance is...
Source: Red Hat Developer Blog
Francisco De Melo Junior, Alexander Barbosa Ayala
2025-10-09 07:00
Red Hat OpenShift now integrates incident detection with OpenShift Lightspeed, enhancing how users analyze and resolve cluster issues. This integration allows for natural language interaction, helping to group related alerts and reduce alert fatigue. Users can easily inquire about incidents, symptoms, and event chains. To set up, ensure you have OpenShift 4.19 and an API key for an LLM provider. Installation steps are available for the Cluster Health MCP server. Explore these new capabilities...
Source: Red Hat Developer Blog
Alberto Falossi, Tomas Remes
2025-10-08 14:16
🚀 Discover how to efficiently manage hundreds to thousands of machine learning models with a systematic approach! This guide outlines a model lifecycle assembly line, focusing on configuration-driven pipelines, version control, and GitOps promotion. Key features include: - Continuous training and versioned pipelines - Data lineage for reproducibility - Safe, automated deployments Learn how to implement these practices in your environment! 🔗 Check out the full details and demo on YouTube!...
Source: Red Hat Developer Blog
clobner
2025-10-08 13:01
🚀 Red Hat Ansible Automation Platform 2.6 is now available! This release enhances automation accessibility with new features. Key updates include a self-service automation portal, the Ansible Lightspeed intelligent assistant, and an on-premise automation dashboard for tracking metrics. Developers can now provide teams with a simple interface to launch automation jobs, visualize their impact, and streamline development environments. Explore the latest version today! #Ansible #Automation...
Source: Red Hat Developer Blog
clobner
2025-10-08 07:16
🌐 Quantum computing is emerging as a transformative technology for developers and businesses. Unlike classical computing, which uses bits (0s and 1s), quantum computing employs qubits, allowing for superposition and entanglement. 🔍 These concepts enable quantum computers to tackle complex problems that classical systems struggle with, such as optimizing financial portfolios or drug design. 🔗 Red Hat is adapting its platforms to integrate quantum capabilities, ensuring developers are prepared...
Source: Red Hat Developer Blog
Maarten Vandeperre
2025-10-07 18:07
🚀 The LLM Compressor 0.8.0 release enhances quantization workflows and extends support for Qwen3 models. Key updates include: 1️⃣ Support for multiple modifiers during a single oneshot compression run, allowing for non-uniform quantization. 2️⃣ Configurable transforms with variable rotation sizes for improved efficiency. 3️⃣ R4 support for SpinQuant-style transforms. 4️⃣ Added quantization for Qwen3 models, including FP8 support. 5️⃣ Improved accuracy for GPTQ W4A16 schemes. Explore the...
Source: Red Hat Developer Blog
Dipika Sikka, Kyle Sayers, Brian Dellabetta
2025-10-07 07:00
Unlock efficient AI inference with llm-d! 🚀 This Kubernetes-native framework introduces KV cache aware routing, reducing latency and improving throughput by directing requests to pods with relevant context in GPU memory. Key features include an External Processing Pod and intelligent routing. With a recent test showing an impressive 87.4% cache hit rate, llm-d enhances performance and optimizes resource use. Learn more about maximizing AI infrastructure efficiency! 📊💡 #AIInference #Kubernetes...
Source: Red Hat Developer Blog
Christopher Nuland
2025-10-07 07:00
🚀 HyperShift revolutionizes Kubernetes management with hosted control planes in Red Hat OpenShift. This innovative feature reduces costs and complexity while enhancing scalability. It allows for faster cluster creation and upgrades, making it easier to manage large fleets. HyperShift’s architecture enables hosted control planes to run on a management cluster, optimizing resource usage. Key considerations include configuring networking, storage, and certificate management for security. Explore...
Source: Red Hat Developer Blog
Andre Rocha
2025-10-07 07:00
🔒 Learn how to sign RPM packages in Red Hat Enterprise Linux 10.1 using quantum-resistant cryptography. This article details the process for developers and vendors to enhance software security through strong signatures. It covers generating OpenPGP keys, configuring RPM, and integrating these steps into existing workflows. Post-quantum cryptography (PQC) aims to protect against future quantum attacks by using hybrid keys and signatures. RPM 6 supports multiple signatures, ensuring both...
Source: Red Hat Developer Blog
Jakub Jelen
2025-10-06 07:00
🚀 Organizations face challenges in running large language models (LLMs) on their infrastructure, especially regarding GPU availability and cost. The Qwen3-Coder-30B-A3B-Instruct model offers strong code-generation capabilities but requires significant GPU resources. To enhance efficiency, quantization is employed to reduce the model's memory footprint while maintaining accuracy. The article outlines a workflow for optimizing and deploying LLMs using Red Hat OpenShift AI, including model...
Source: Red Hat Developer Blog
Philip Hayes