MapR Data Fabric for Kubernetes provides persistent storage for containers and enables the deployment of stateful containers. It addresses the limitations of container use by providing easy and full data access from within and across clouds and on-premises deployments. Now stateful applications can easily be deployed in containers for production use cases, machine learning pipelines, and multi-tenant use cases.
As organizations adopt containers at a larger scale, and as they move containers as a deployment model to production, other aspects such as storage, monitoring, and performance becomes essential as well. MapR Data Fabric for Kubernetes provides several features that will assist organizations in their journey with containers.
MapR integration with the Kubernetes storage plugin allows for MapR volumes to be mounted for access by containers, through the CSI model APIs. As new containers are deployed, data volumes can be created and retained, even when containers are deleted. If containers are moved across environments, the MapR global namespace provides access to the data, independent of where the containers reside, using the Kubernetes primitives such as PersistentVolumes, PersistentVolumeClaims, StorageClasses.
Applications can be synchronized and updated with a unified view and without disruption. Data stored in MapR can benefit from the full-fledged advantages of high availability, security, data protection, and disaster recovery.
By using Kubernetes DaemonSet on MapR, containers can scale seamlessly without any disruption to the existing environment. With its “grow as you go” feature, MapR handles growth in data without it having to be moved to a separate, dedicated environment. Containers deployed in cloud or in edge clusters can easily gain access to data in MapR, and containers can be seamlessly added and moved across clusters.
Containers running on client hosts can achieve high bandwidth access to data hosted on MapR. Container clusters managed by Kubernetes can provision and access data on any type of hardware infrastructure: SSD, NVMe, or HDD. As containers and data scale, performance scales linearly.
MapR tickets can be used to set up user authentication and authorization methods to access the persistent data. As more containers are added, each gets its own secure, persistent sandbox of compute, storage, and memory resources. This is an essential benefit, particularly for organizations that have multi-tenant environments.
MapR distributed architecture ensures critical data is never lost, through tunable levels of replications and distributing copies across the cluster. Automatic failover ensures data is always available without any disruption to containerized applications.
Enterprise data services, like mirroring and consistent point-in- time snapshots, protect critical data. Applications are becoming fluid because of containers, since containers can be moved or recreated across clusters. Recovering from major outages or having planned outages is made easier by migrating containers and mirroring/snapshotting data for containers.
MapR Data Fabric for Kubernetes, combined with the capabilities of the MapR Converged Data Platform, is beneficial to organizations desiring to run workloads in containers.