AI/ML
Artificial intelligence and machine learning workloads that process large datasets through compute-intensive training and inference, and typically depend on scalable compute (often GPUs) and reliable data pipelines.
Artificial intelligence and machine learning workloads that process large datasets through compute-intensive training and inference, and typically depend on scalable compute (often GPUs) and reliable data pipelines.
The process of updating legacy applications and infrastructure to cloud-native architectures such as containers, microservices, and orchestration, often by re-platforming or refactoring monoliths rather than rewriting them.
A copy of data and application state stored separately so systems can be restored after loss, corruption, ransomware, or failure; in Kubernetes it should capture both persistent data and application configuration/metadata.
Business continuity and disaster recovery treated together as one resilience strategy—continuity keeps critical operations running during disruption, while disaster recovery restores systems and data after a major outage.
Storage presented as raw, fixed-size blocks that an application or file system formats and manages directly, offering the low latency and high performance favored by databases and transactional workloads.
An organization's ability to keep critical operations and services running during and after a disruption, supported by planning, redundancy, and defined recovery procedures.
Continuous integration and continuous delivery/deployment—a pipeline approach that automates building, testing, and releasing software so changes reach production faster and more reliably with less manual effort.
Storage designed for dynamic, containerized, orchestrated environments such as Kubernetes, adding capabilities like dynamic provisioning, scalability, and workload portability that traditional storage lacks.
A group of machines (nodes) managed together by Kubernetes to run containerized workloads, made up of a control plane that coordinates the cluster and worker nodes that run the applications.
A Kubernetes object for storing non-sensitive configuration as key-value pairs, decoupling settings from container images so applications can consume configuration without being rebuilt.
A lightweight, isolated package that bundles an application with its dependencies so it runs consistently across environments, sharing the host OS kernel rather than requiring a full guest operating system.
A KubeVirt component that imports, uploads, and clones virtual machine disk images into Kubernetes storage (PVCs), enabling VMs to run alongside containers.
The automated scheduling, deployment, scaling, networking, and lifecycle management of containers across many machines to keep workloads healthy and available; Kubernetes is the most widely used example.
A vendor-neutral standard interface that lets Kubernetes and other orchestrators integrate with many storage systems through drivers, without storage-specific logic being baked into the orchestrator's core.
The set of Kubernetes components—such as the API server, scheduler, and controllers—that make global decisions about the cluster and continuously reconcile its actual state toward the desired state.
A storage plugin that implements the Container Storage Interface so Kubernetes can dynamically provision, attach, mount, and manage storage from a specific backend or vendor.
An extension mechanism that adds new, custom resource types to the Kubernetes API, letting users manage application-specific objects with the same tooling used for built-in resources.
A Kubernetes controller that ensures a copy of a specific pod runs on every selected node, commonly used for node-level agents such as logging, monitoring, or storage daemons.
A managed database delivery model that abstracts provisioning, scaling, patching, backups, and maintenance so teams can consume databases on demand without managing the underlying infrastructure.
A Kubernetes controller for stateless applications that maintains a desired number of pod replicas and handles rolling updates and rollbacks, typically with no downtime.
A culture and operating model that combines development and operations through automation, shared ownership, and continuous feedback to deliver software faster and more reliably.
A backup that copies all data changed since the last full backup; it grows larger over time but restores faster than incremental backups because only the full plus one differential are needed.
The processes, tools, and plans used to restore applications and data after a serious outage or disaster, typically measured against recovery point and recovery time objectives (RPO/RTO).
Training a machine learning model across multiple workers, GPUs, or nodes in parallel to cut training time and handle datasets or models too large for a single machine.
A popular container platform and toolchain used to build, package, and run container images, and a key driver of the container adoption that Kubernetes orchestrates at scale.
Temporary storage tied to the lifecycle of a pod or container that is deleted when the workload stops or restarts, making it suitable for caches, scratch space, and other non-persistent data.
Automatically switching workloads to a standby system, node, or site when the primary becomes unavailable, minimizing downtime and preserving service availability.
Storage accessed through a shared file system interface and organized as a hierarchy of files and directories, often shared concurrently by multiple clients or pods (for example, over NFS).
A backup that copies the entire selected dataset regardless of prior backups, providing the simplest and most complete restore point at the cost of more time and storage.
Scheduling logic that places workloads on nodes with the required graphics processing units available, ensuring GPU-intensive jobs such as AI/ML training run where they can be accelerated.
The software layer that creates and manages virtual machines by abstracting and allocating physical hardware, allowing multiple isolated guest operating systems to share one host.
A backup that copies only the data changed since the last backup of any type, minimizing backup time and storage but requiring the full plus all subsequent increments to restore.
An open-source toolkit for building, training, deploying, and operating machine learning workflows on Kubernetes, providing portable, scalable ML pipelines across environments.
An open-source platform for deploying, scaling, and managing containerized applications across a cluster, automating scheduling, self-healing, load balancing, and rollouts.
An open-source Kubernetes extension that lets virtual machines run and be managed through Kubernetes APIs, so VMs and containers can be operated side by side on one platform.
Moving a running virtual machine from one host to another with little or no disruption to the workload, used for maintenance, load balancing, and avoiding downtime.
The operational practices and tooling used to deploy, monitor, version, retrain, and govern machine learning systems in production, applying DevOps principles to the ML lifecycle.
A document-oriented NoSQL database that stores data as flexible JSON-like documents and is commonly deployed in distributed, cloud-native environments for horizontal scalability.
A widely used open-source relational database management system that stores structured data in tables and is commonly used to run transactional application workloads.
A physical or virtual machine that Kubernetes uses to host and run pods, providing the compute, memory, and storage resources the scheduler assigns workloads to.
Rules that constrain where a pod or volume can be placed based on node labels or topology, used to steer workloads onto nodes that meet specific hardware or location requirements.
A high-performance storage protocol built for flash/SSD devices that uses parallel queues over PCIe to cut latency and sharply increase throughput compared with older protocols like SATA.
Storage that manages data as discrete objects bundled with metadata and a unique identifier, accessed through an API (such as S3) rather than a file system, and well suited to large-scale unstructured data.
Red Hat's enterprise Kubernetes platform that layers integrated security, developer, CI/CD, and operations tooling on top of upstream Kubernetes for production use.
Red Hat's supported, KubeVirt-based capability that runs and manages virtual machines alongside containers within OpenShift using the same Kubernetes tooling.
A Kubernetes-native software pattern that uses custom resources and controllers to automate complex application lifecycle tasks—such as deployment, upgrades, and backups—by encoding operational knowledge into software.
A design approach that extends the Kubernetes API with custom resources and controllers to automatically manage stateful or complex systems, continuously reconciling them toward their desired state.
Storage that retains data beyond the life of any individual container or pod restart, essential for stateful applications such as databases that must not lose data.
A cluster-level Kubernetes object representing a piece of provisioned storage, decoupled from any single pod so its data persists independently of workload lifecycle.
A request for storage by a workload—specifying size and access mode—that Kubernetes binds to a matching Persistent Volume, serving as the abstraction pods use to consume storage.
The practice of building and maintaining internal platforms and self-service tooling that abstract infrastructure complexity, improving developer productivity and consistency in software delivery.
The smallest deployable unit in Kubernetes, wrapping one or more tightly coupled containers that share network and storage and are scheduled together on the same node.
An open-source Kubernetes management platform used to deploy, operate, and govern multiple clusters across on-premises, cloud, and edge environments from a single interface.
A Kubernetes volume access mode allowing a volume to be mounted read-write by many nodes simultaneously, enabling shared storage for workloads that need concurrent access.
A Kubernetes volume access mode allowing a volume to be mounted read-write by a single node at a time, typical for block storage used by one pod.
The maximum acceptable amount of data loss, expressed as a time window (for example, 15 minutes), which dictates how frequently backups or replication must occur.
The target maximum time to restore systems or services after an outage, used to guide the design of recovery processes and infrastructure.
Maintaining multiple synchronized copies of data across nodes or sites so workloads remain available and can recover quickly from hardware failure or disaster.
A Kubernetes authorization model that grants permissions through roles bound to users or service accounts, controlling who can perform which actions on which resources.
The process Kubernetes uses to decide which node should run a given pod, weighing resource requests, constraints, affinity rules, and policies.
A Kubernetes object for storing and distributing sensitive data such as passwords, tokens, and keys, keeping it separate from application code and optionally encrypted at rest.
A point-in-time copy of a volume or dataset that captures its state for backup, cloning, rollback, or recovery, and is typically created quickly without copying all the data.
Solid-state storage that uses flash memory instead of moving parts, delivering faster access, higher throughput, and lower latency than traditional spinning hard disk drives.
A Kubernetes controller for stateful applications that gives pods stable, unique network identities, persistent storage, and ordered, predictable deployment and scaling.
A Kubernetes resource that defines a class of storage—its provisioner and parameters—so Persistent Volumes can be dynamically provisioned on demand with the desired performance and behavior.
SUSE's virtualization platform (Harvester) for running and managing virtual machine workloads using a Kubernetes-centric, hyperconverged model that unifies compute, storage, and networking.
A Kubernetes mechanism in which taints on a node repel pods unless those pods carry a matching toleration, used to reserve nodes for specific workloads.
The use of an abstraction layer to run multiple isolated operating environments on shared physical hardware, improving utilization, isolation, and flexibility.
A software-defined computer that emulates hardware and runs its own guest operating system, isolated from other VMs on the same physical host by a hypervisor.
The KubeVirt runtime object representing a single running instance of a virtual machine, analogous to how a pod represents a running container workload.
VMware's live-migration technology for moving running virtual machines between physical hosts with no downtime, used for maintenance and load balancing.
A storage resource mounted into a container or pod that lets applications read and write data, and—depending on its type—can outlive container restarts to preserve data.
A machine in a Kubernetes cluster that runs application workloads (pods) along with supporting services such as the kubelet and container runtime, under the direction of the control plane.
No terms match your search.