Articles from Source: Red-Hat-Developer-Blog

How to run a Red Hat-powered local AI audio transcription

2026-03-25 07:01
Unlock the potential of AI with local audio transcription! 🎀 This guide walks you through setting up a transcription application using Red Hat AI. With just a few Python commands, you can keep your audio files secure and offline. Learn how to install the necessary tools, download the Red Hat model, and transcribe audio files seamlessly. Dive into the world of open-source AI and enhance your projects! πŸš€ #RedHatAI #AudioTranscription #OpenSource #AIFuture #PythonProgramming
Source: Red Hat Developer Blog
Seth Kenlon

Run Model-as-a-Service for multiple LLMs on OpenShift

2026-03-24 07:16
Unlock the power of multiple LLMs with Model-as-a-Service (MaaS) on OpenShift! πŸš€ This guide details how to create a unified entry point for AI inference, simplifying developer interactions with various models like Qwen and TinyLlama. Learn to build intelligent routing that efficiently directs traffic and reduces GPU waste. One endpoint means unified authentication and simplified monitoring. Explore the integration of llm-d capabilities into Red Hat OpenShift AI for a streamlined enterprise...
Source: Red Hat Developer Blog
Vladimir Belousov

Evaluate OpenShift cluster health with the cluster observability operator

2026-03-24 07:16
🌐 Red Hat OpenShift introduces a new component health overview within the cluster observability operator, now in Developer Preview. This feature aids in assessing the health of the OpenShift control plane and other components, categorizing their status as OK, warning, or error through a Perses dashboard. To use, install cluster observability operator 1.4 or later via OperatorHub and enable recommended monitoring. Explore Prometheus metrics to view component relationships and health statuses....
Source: Red Hat Developer Blog
Tomas Remes

Integrate Red Hat Advanced Cluster Management with Argo CD

2026-03-24 07:16
🌐 Red Hat Advanced Cluster Management integrates with Argo CD for enhanced Kubernetes application management. This setup allows clusters to subscribe to Git repositories using Channels and Subscriptions, supporting both push and pull models. πŸ” Key features include the ApplicationSet Custom Resource and the ability to configure sync policies for better resource management. πŸ“Š Audit logs play a crucial role in tracking actions and troubleshooting issues. Always verify the Argo CD console and...
Source: Red Hat Developer Blog
Francisco De Melo Junior

Upgrade Advanced Cluster Management hubs without disruption

2026-03-23 07:00
Upgrading Red Hat Advanced Cluster Management hubs can be challenging due to risks like downtime and upgrade failures. The new solution, managed cluster migration, allows for parallel hub deployment. This means a new hub version can be set up while the old one remains operational, ensuring zero disruption during the upgrade process. The migration is monitored closely, with automatic rollbacks in case of failures, making the upgrade process safer and more reliable. #RedHat #Kubernetes...
Source: Red Hat Developer Blog
Dang Peng Liu

Eval-driven development: Build and evaluate reliable AI agents

2026-03-23 07:00
πŸš€ Check out insights from our journey in developing the rh-ai-quickstart/it-self-service-agent! We explored an evaluations framework tailored for AI agents, emphasizing the need for comprehensive testing due to their inherent variability. Key stages of our evaluation journey include: 1️⃣ Manual testing with predefined conversations 2️⃣ Automated evaluations with custom metrics 3️⃣ Continuous integration for ongoing improvements Learn more about how we integrated these practices into our...
Source: Red Hat Developer Blog
Michael Dawson

Hybrid loan-decisioning with OpenShift AI and Vertex AI

2026-03-19 07:01
Explore how modern financial applications make loan decisions using hybrid machine learning systems! πŸ€–πŸ’Ό The architecture features a loan approval classifier on Google Cloud with Vertex AI, while an ONNX regression model for interest rate prediction runs on Red Hat OpenShift AI on-premise. This setup ensures sensitive data remains secure and compliant. πŸ”’ Key components include a lightweight React frontend and a Llama-based chatbot for enhanced user interaction, all orchestrated seamlessly....
Source: Red Hat Developer Blog
Harshil Sabhnani

Rebalance hub workloads with managed cluster migration

2026-03-19 07:01
🌐 As Kubernetes deployments grow, managing workloads efficiently becomes crucial. Red Hat Advanced Cluster Management introduces a multicluster global hub for dynamic redistribution of managed clusters. This system allows for workload balancing, easing the strain on individual hubs and reducing latency. To implement this, ensure the global hub is set up and source clusters are imported correctly. This approach supports various scenarios, including capacity redistribution and geographic...
Source: Red Hat Developer Blog
Dang Peng Liu

How to operate OpenShift in air-gapped environments

2026-03-19 07:01
Operating Red Hat OpenShift in air-gapped environments enhances security by eliminating direct internet access. Organizations must manage all software and updates through controlled processes to meet strict compliance and sovereignty requirements. πŸ”’πŸŒ Key challenges include initial setup and ongoing maintenance. It’s crucial to establish a defined operational cadence for mirroring platform updates, managing operator catalogs, and ensuring security fixes. πŸ“…πŸ”„ Implementing a mirror factory can...
Source: Red Hat Developer Blog
Phillip Knezevich

Automate test and failure analysis via streams for Apache Kafka

2026-03-19 07:01
In enterprise software testing, syncing failure analysis between ReportPortal and Polarion is crucial. This article outlines an event-driven solution using Apache Kafka, Debezium CDC, and Quarkus to automate this sync. The approach ensures real-time updates with minimal lag, eliminating manual data entry. Key challenges included handling divergent timelines and ensuring consistent data flow across platforms. The CDC method captures updates at the database level, allowing for seamless...
Source: Red Hat Developer Blog
Guannan Sun

LLM Compressor v0.10: Faster compression with distributed GPTQ

2026-03-18 15:21
πŸš€ Exciting news for AI developers! LLM Compressor v0.10 has launched, introducing faster compression for large language models (LLMs). Key features include: - **Distributed GPTQ**: Achieve up to 3.8x speedup using multiple GPUs. - **Compressed-tensors offloading**: Compress models beyond your memory capacity. - **GPTQ FP4 microscale support**: Utilize NVFP4 and MXFP4 quantization schemes. This update enhances performance and accuracy, making it a valuable tool for managing large models...
Source: Red Hat Developer Blog
Kyle Sayers, Charles Hernandez, Dipika Sikka

How Advanced Cluster Management simplifies rule management

2026-03-18 07:01
Managing security for secondary networks can be complex and time-consuming. The article discusses how Red Hat Advanced Cluster Management for Kubernetes simplifies this task. By using ConfigMaps, you can define network rules centrally on a hub cluster. This automation allows for the creation of localized MultiNetworkPolicies across all managed clusters, reducing manual efforts and errors. The system ensures compliance and security posture at scale, making it easier for teams to manage their...
Source: Red Hat Developer Blog
Moyo Oyegunle

Prepare to enable Linux pressure stall information on Red Hat OpenShift

2026-03-18 03:01
πŸš€ Red Hat OpenShift 4.21 introduces Linux pressure stall information (PSI) via MachineConfig, enhancing resource monitoring for CPU, memory, and I/O. Enabling PSI helps identify resource contention and hidden bottlenecks, improving autoscaling and debugging. However, it increases memory usage for Prometheus pods, showing a 42% rise with over 500 test containers. For more insights on PSI metrics and their impact, check out the article. #OpenShift #Linux #PSI #Kubernetes #CloudComputing
Source: Red Hat Developer Blog
Qiujie Li

Advanced Cluster Management 2.16 right-sizing recommendation GA

2026-03-17 07:00
πŸ” Red Hat Advanced Cluster Management 2.16 has launched its right-sizing recommendations for namespaces and OpenShift Virtualization workloads, now available for enterprise use. πŸ“Š This feature provides a unified approach to resource optimization, helping organizations manage clusters efficiently. Key benefits include improved resource insight, data-driven recommendations, and insightful visualizations through integrated Grafana dashboards. πŸ’‘ Administrators can identify over and under-...
Source: Red Hat Developer Blog
Darshan Vandra, Raj Zalavadia

Configure NVIDIA Blackwell GPUs for Red Hat AI workloads

2026-03-16 20:30
πŸ“’ Exciting news for Red Hat AI users! The NVIDIA RTX PRO 4500 Blackwell Server Edition offers enhanced GPU acceleration for enterprise data centers. This server edition boosts performance for various Red Hat AI applications, making it easier to build and deploy AI workloads efficiently. To configure, install the NVIDIA GPU Operator in Red Hat OpenShift, ensuring optimal settings for driver and kernel modules. Discover the benefits of compact, power-efficient AI deployments with Blackwell!...
Source: Red Hat Developer Blog
Erwan Gallen, Tarun Kumar, Antonin Stefanutti, Selbi Nuryyeva, Michey Mehta

Unlocking UBI to Red Hat Enterprise Linux container images

2026-03-16 12:30
Unlocking UBI for Red Hat Enterprise Linux (RHEL) enhances your container development. While the Red Hat Universal Base Image (UBI) is lightweight, it lacks some essential packages. By utilizing a no-cost Red Hat Developer subscription, you gain access to a broader set of RHEL packages, including databases and additional tools. This subscription supports both individual projects and business workloads, ensuring you have a reliable foundation for your development needs. Explore the benefits of...
Source: Red Hat Developer Blog
Louis Imershein

Zero trust GitOps: Build a secure, secretless GitOps pipeline

2026-03-13 07:01
πŸ” Discover how OpenShift GitOps enhances security with short-lived tokens! This integration with the external secrets operator allows for secure management of credentials, minimizing the risk of breaches. Short-lived tokens provide limited access and ensure continuous authentication. Learn more about this innovative approach to secure GitOps pipelines! #OpenShift #GitOps #Cybersecurity #DevOps #Kubernetes
Source: Red Hat Developer Blog
Nick Png

How to manage Red Hat OpenShift AI dependencies with Kustomize and Argo CD

2026-03-13 07:01
πŸš€ Managing dependencies for Red Hat OpenShift AI can be complex, but the new odh-gitops repository simplifies the process. This repository offers a GitOps-ready template with Kustomize manifests for deploying all necessary dependencies. You can easily apply configurations using the `oc apply -k` command or through Argo CD for automated deployment. For those preferring Helm, a chart alternative is in development and will be available soon. Check out the odh-gitops repository to learn more!...
Source: Red Hat Developer Blog
Davide Bianchi, Andrea Tarocchi

How to develop agentic workflows in a CI pipeline with cicaddy

2026-03-12 07:00
πŸš€ Discover how to develop agentic workflows in your CI pipeline using cicaddy! Traditional agentic platforms can be complex, but cicaddy simplifies this by allowing you to integrate AI directly within your existing CI/CD workflows. This approach enhances your workflows without replacing them, using Large Language Models for tasks like report generation and anomaly detection. Learn how to automate tasks seamlessly! πŸ”πŸ”„ #CICD #AIWorkflow #DevOps #Automation #Cicaddy
Source: Red Hat Developer Blog
Guannan Sun

Accelerated expert-parallel distributed tuning in Red Hat OpenShift AI

2026-03-11 15:50
Red Hat OpenShift AI enhances AI performance through distributed fine-tuning of foundation models. The article discusses challenges in coordinating computation and communication across GPUs. To address this, it introduces the open-source library, fms-hf-tuning, which supports efficient fine-tuning of language and vision-language models. Key features include data preprocessing, throughput optimization, and expert parallelism techniques. The library aims to improve memory efficiency and...
Source: Red Hat Developer Blog
Karel Suta, Amita Sharma

Improve code quality and security with PatchPatrol

2026-03-11 07:01
πŸš€ Introducing PatchPatrol, an open-source tool designed to enhance code quality and security for enterprise teams using Red Hat OpenShift. PatchPatrol integrates AI-powered analysis directly into your CI/CD pipelines, ensuring every code change meets security standards. It offers dual modes: one for code quality and another focused on security vulnerabilities, including OWASP Top 10 detection. With flexible backend options, teams can choose between local and cloud models based on their needs....
Source: Red Hat Developer Blog
Herve Beraud

Agent Skills: Explore security threats and controls

2026-03-10 07:16
πŸ“’ Anthropic has launched Agent Skills, a new functionality now available across various agents, including Goose. This feature allows agents to perform tasks tailored to user needs using structured skills organized in folders. πŸ” The article discusses the importance of managing security threats and access controls with Agent Skills. Key considerations include proper folder permissions, vulnerability management, and the risks associated with executable scripts. πŸ”’ To mitigate potential risks like...
Source: Red Hat Developer Blog
Florencio Cano Gabarda

How to run Slurm workloads on OpenShift with Slinky operator

2026-03-10 07:16
πŸš€ Exciting advancements in high-performance computing! This article covers how to run Slurm workloads on OpenShift using the Slinky operator. Slurm is a robust workload manager, and combining it with OpenShift enhances scalability and automation for HPC environments. πŸ”§ The Slinky operator simplifies deployment, scaling, and lifecycle management of Slurm components within Kubernetes. This integration allows teams to utilize familiar tools while leveraging containerized infrastructure. πŸ’‘ Key...
Source: Red Hat Developer Blog
Prudhvi Vuda, Swati Kale

Effortless Red Hat Enterprise Linux virtual machines with Libvirt and Kickstart

2026-03-10 03:01
Creating virtual machines (VMs) for development is essential for testing and reproducing production environments. This article outlines a straightforward method using Libvirt and Kickstart scripts in Red Hat Enterprise Linux (RHEL) to automate VM provisioning. With Kickstart, developers can set up customized test VMs without needing extensive network infrastructure. Both Windows and MacOS users can utilize these scripts effectively, even within WSL2 or through Podman Desktop. Learn how to...
Source: Red Hat Developer Blog
Fernando Lozano

5 steps to triage vLLM performance

2026-03-09 14:12
Navigating LLM performance in production can be challenging. The article outlines a 5-step diagnostic workflow for optimizing vLLM deployments. πŸ” Start by defining your performance objectives based on workload profiles: throughput-sensitive, latency-sensitive, or bursty. 1️⃣ Identify where latency occurs by analyzing Time to First Token (TTFT) and Inter-Token Latency (ITL). 2️⃣ Monitor server saturation to understand request queue dynamics. 3️⃣ Evaluate GPU memory and KV cache health to...
Source: Red Hat Developer Blog
David Whyte-Gray, Thameem Abbas Ibrahim Bathusha, Michael Goin, Ashish Kamra

Automate AI agents with the Responses API in Llama Stack

2026-03-09 14:12
πŸš€ Automate AI agents with the Responses API in Llama Stack! This article discusses how the Responses API enhances AI agent orchestration while maintaining precise control over conversations. It automates tool calls and state management, facilitating smoother interactions. Learn about the benefits of adopting this API, especially for IT process automation, and explore hands-on examples through the AI quickstart series. For more insights, check out the full article! πŸ“ˆπŸ€– #AI #Automation...
Source: Red Hat Developer Blog
Michael Dawson

Smarter multi-cluster scheduling with dynamic scoring framework

2026-03-09 03:01
πŸš€ In multi-cluster management, effective workload deployment is crucial. The Placement API and PlacementScores from Open Cluster Management enable dynamic cluster selection based on various metrics. πŸ“Š The new Dynamic Scoring Framework automates cluster scoring using Prometheus metrics, making real-time decisions easier. It simplifies the integration process, allowing for tailored scoring logic. πŸ”§ Developers can create custom scorers for cost efficiency, predictive metrics, and more, enhancing...
Source: Red Hat Developer Blog
Jian Qiu

What's new in network observability 1.11

2026-03-06 08:01
πŸš€ Network Observability 1.11 is here! This release aligns with Red Hat OpenShift Container Platform 4.21 and remains backward-compatible. Key features include: - A new Service deployment model for better resource management. - Introduction of FlowCollectorSlice CRD for namespace-level control. - Zero-click Loki setup for non-production use. - Enhanced DNS tracking and network health improvements. Explore more in the OpenShift web console! πŸŒπŸ” #NetworkObservability #RedHat #OpenShift...
Source: Red Hat Developer Blog
Steven Lee

From local prototype to enterprise production: Private speech transcription with Whisper and Red Hat AI

2026-03-06 07:00
Discover how to run OpenAI's Whisper model locally on Apple Silicon using vLLM. This setup allows for private audio transcription, crucial for sensitive environments like healthcare and finance. The article outlines various use cases, including AI coaching, compliance monitoring, and secure meeting transcriptions. To transition from local development to enterprise production, Red Hat AI Inference Server offers a robust solution. πŸ”’πŸ’»πŸ“Š #AI #Transcription #OpenAI #DataPrivacy #RedHat
Source: Red Hat Developer Blog
Carlos Condado, Yuchen Fama

Temurin JDK 25 now available in Red Hat Customer Portal

2026-03-06 07:00
πŸš€ Great news for Java developers! Temurin JDK 25 for x64 Windows is now available in the Red Hat Customer Portal, offering a production-ready option for running Java. Red Hat supports Temurin builds, and with this release, developers can easily access it. There will be no Red Hat build of OpenJDK for JDK 25 on 64-bit Windows, making Temurin the primary choice. πŸ” This flexibility allows teams to standardize tools while selecting the best support options. Red Hat ensures that Temurin behaves...
Source: Red Hat Developer Blog
Jeff Beck

How to scale enterprise federated AI with Flower and OCM

2026-03-05 08:01
🌐 Federated AI transforms traditional machine learning by bringing models to the data instead of vice versa. This method ensures local training on distributed nodes, keeping raw data secure. πŸ’‘ Flower is the leading open-source framework for federated AI, widely adopted by tech giants and institutions for its simplicity and versatility across ML frameworks. πŸ”— Learn how Flower integrates with Open Cluster Management (OCM) to streamline deployment and enhance privacy compliance. #FederatedAI...
Source: Red Hat Developer Blog
Meng Yan

Boring RAG: When similarity is just a SQL query

2026-03-05 07:00
Retrieval-augmented generation (RAG) is a method for answering questions using your own content without relying on general LLMs. It follows a simple pattern: retrieve context, then answer. This article explores a straightforward RAG implementation with Apache Camel and PostgreSQL, focusing on making the process easy to understand and debug. Key steps include indexing content, retrieving information, and providing answers based on the context. Learn about embeddings, chunking, and how to...
Source: Red Hat Developer Blog
Ivo Bek

How to collaborate with AI to improve your Ansible skills

2026-03-04 08:00
Discover how to enhance your Ansible skills with AI! πŸ€– In a recent article, the author shares their journey of using AI to create a reporting playbook for auditing Red Hat Enterprise Linux (RHEL) versions. Key insights include handling unreachable hosts and utilizing Jinja2 for efficient data grouping. The final playbook generates a clear, formatted report, simplifying server management. Learn more about this collaboration and the techniques used! πŸ“ŠπŸ’» #Ansible #AI #SysAdmin #RedHat #Automation
Source: Red Hat Developer Blog
Roberto Nozaki

Estimate GPU memory for LLM fine-tuning with Red Hat AI

2026-03-04 07:15
Unlock the potential of fine-tuning language models with Red Hat AI's Training Hub! πŸš€ This open-source Python package helps you customize pre-trained models on your datasets. However, fine-tuning demands more GPU memory than inference. To avoid costly errors, the new memory estimator tool calculates your memory needs efficiently. Learn how to optimize your GPU setup, adjust training components, and streamline your fine-tuning process. #RedHatAI #MachineLearning #FineTuning #AI #OpenSource
Source: Red Hat Developer Blog
Mohib Azam

Serve and benchmark Prithvi models with vLLM on OpenShift

2026-03-03 15:00
πŸš€ Dive into the world of AI with the latest article on serving and benchmarking Prithvi models using vLLM on OpenShift! The article outlines two main activities: deploying the Prithvi model with a traditional setup and utilizing KServe for dynamic scaling. Key prerequisites include a Red Hat OpenShift cluster with NVIDIA GPU support. Interested in testing your own Earth and space models? The guide provides detailed steps to get started! 🌌 #AI #OpenShift #MachineLearning #KServe #RedHat
Source: Red Hat Developer Blog
Michele Gazzetti, Michael Johnston, Christian Pinto, Erwan Gallen

Optimize PyTorch training with the autograd engine

2026-03-03 13:47
Discover the power of the PyTorch autograd engine! πŸ” This article explores how autograd calculates gradients, builds computational graphs, and manages memory efficiently during backpropagation. Understanding these concepts can enhance your deep learning models. Key points include: - Automatic differentiation with dynamic graph construction. - Memory optimization through pruning unnecessary computations. - The significance of forward and backward passes. #DeepLearning #PyTorch #Autograd...
Source: Red Hat Developer Blog
Vishal Goyal

Practical strategies for vLLM performance tuning

2026-03-03 07:01
πŸš€ Tuning large language models (LLMs) like vLLM involves balancing hardware, workload, and user experience. Start with a realistic test dataset to ensure accurate performance optimization. Tools like GuideLLM can help create benchmarks that reflect actual usage patterns. Key considerations include input/output shapes and memory utilization. Adjust GPU configurations to find the optimal balance for your specific needs. For more insights, check the full article! πŸ“ŠπŸ’» #vLLM #PerformanceTuning...
Source: Red Hat Developer Blog
Trevor Royer

What’s new in Ansible Certified Content Collection for AWS

2026-03-02 08:01
πŸš€ The release of amazon.aws 11.0.0 enhances AWS automation by prioritizing stability and supportability. Key updates include improved debugging for Amazon S3, standardized module behavior, and the removal of deprecated features. Users are encouraged to upgrade to benefit from better performance and consistency. Stay informed and streamline your cloud management! #AWS #Ansible #CloudAutomation #RedHat #DevOps
Source: Red Hat Developer Blog
Alina Buzachis

How to automate Ceph RGW user management on OpenShift with GitOps

2026-03-02 08:01
Unlock the potential of automated user management in Ceph RGW on OpenShift with GitOps! πŸ“¦βœ¨ This guide provides a workflow using OpenShift GitOps and the external secrets operator for efficient object storage quota enforcement. Learn how to automate user account creation and manage quotas seamlessly. Key prerequisites include OpenShift Data Foundation, rook-ceph operator, and access to a Git repository. Explore how to bridge traditional storage management with modern GitOps practices! #Ceph...
Source: Red Hat Developer Blog
Mohammad Ahmad, James Blair

GDAL 3.4 package: Full-featured GIS functionality on RHEL

2026-02-27 08:01
When using RHEL 9 for geospatial tasks, the GDAL package offers only basic functionality. Common operations may fail due to limited driver support, which is intentional to enhance security and stability. 🌍 For full GDAL capabilities, the gdal3.4 package is available in EPEL, allowing parallel installation without causing file conflicts. This provides access to essential drivers for formats like GeoTIFF and GeoJSON. πŸ—ΊοΈ Key features include prefixed utilities to avoid conflicts and...
Source: Red Hat Developer Blog
Filip Janus