Red Hat 블로그
Hyperconvergence is a key topic in IT planning across industries today. As customers look to lower costs and simplify day to day management of their IT operations, the hyperconverged model emerges as fit in a number of operational use cases.
Convergence began at the hardware level, with compute, network, and storage appearing in consolidated platforms, but it’s now accelerating as hyperconvergence goes “software defined”. As a leading software infrastructure stack provider, Red Hat recognizes that reducing the overall moving parts in your infrastructure and simplifying the procurement and deployment processes are core requirements of the next generation elastic datacenter.
Applying a solutions-aligned lens, Red Hat is innovating software defined compute-storage solutions across the portfolio, designed to meet the needs of a broad customer base with diverse requirements. As a vendor-partner in this journey, we recognize the value of bringing storage close to your compute and eliminating the need for discreet storage tier. Doing so across both traditional virtualization and cloud, as well as containers and leveraging our industry-proven software defined storage assets - Red Hat Gluster and Red Hat Ceph Storage - we’ve defined a robust set of efficient, solution-aligned hyperconverged offerings.
This blog provides a short overview of several areas where we see hyperconverged software defined architectures aligning with use cases, with a focus on
software defined storage.
Container-native Storage Provisioning & Management
With container technologies featuring prominently in next generation data center planning, many IT professionals are focusing on Red Hat OpenShift Container Platform, which provides users with a simple, elegant system to manage, deploy, and provision their containerized applications at a massive scale. Red Hat is now enhancing the platform with persistent storage capabilities, integrating the distributed Red Hat Gluster Storage offering for reliable, shared storage for their critical applications.
To that end, Red Hat recently announced container-native storage for OpenShift Container Platform. Prior to the container-native storage enhancements, OpenShift deployments depended on separate storage systems exclusively - requiring separate management and hardware resources.
With this capability, administrators can now deploy their storage solution onto existing OpenShift nodes not only to achieve a high level of resource utilization, but also to manage and provision storage using OpenShift services. Container-native storage for OpenShift Container Platform is one way Red Hat is revolutionizing the way we think about storage deployment, and management.
Infrastructure Consolidation - Virtualization & Software-defined Storage
As growing compute and storage demands put a strain on remote office operations, increasingly we find IT teams looking for efficient models to scale compute and storage capacity without having to continue to scale IT staff at these remote facilities. Additionally, space and cooling tend to be at a premium in these use cases and hence continuing to stand-up what are essentially smaller versions of the infrastructure found in the central data centers does not scale.
Hyperconverged virtualization and software-defined storage provides an ideal infrastructure model for scaling these operations with a reduced burden on acquisition, deployment, footprint, and in day to day management. To meet this cross-industry need, Red Hat has been building on the existing functional relationship between Red Hat Enterprise Virtualization and Red Hat Gluster Storage, furthering the existing affinity between these two robust, industry-proven systems by converting them into a single, unified cluster configuration.
Extending the existing Red Hat Enterprise Virtualization and Red Hat Gluster Storage integration to hyperconverged is realized in part via additional workflow optimizations - deepening and extending them across compute and storage as a unit - as well as converging the hardware and software monitoring and fault management handling and reporting. This results in streamlined operation via a single management control plane across compute and storage, and lays the foundation for advanced orchestration and automation capabilities via integration with the broader Red Hat infrastructure stack.
We have now introduced a limited availability roll out of Red Hat Enterprise Virtualization and Red Hat Gluster Storage. More information is available via your Red Hat representative.
Scalable Storage for Infrastructure-as-a-Service Deployments
Just as compute and storage converge in the virtualization arena, so too are software-defined storage and private cloud solutions. Customers recognize that their OpenStack private clouds require software-defined storage resources in order to scale fluidly, and further that Red Hat Ceph Storage is today overwhelmingly preferred in these deployments. Because it offers a single, flexible platform supporting block storage - both persistent and ephemeral - as well as object and file storage, we believe Red Hat Ceph Storage delivers the ideal storage platform for your OpenStack cloud.
A set of use cases are emerging in which hyperconverged Ceph with OpenStack provides an optimized deployment model. Among these are NFV (Network Functions Virtualization) architectures. While OpenStack and Ceph are often associated with building large scale out clouds, NFV deployments tend to be small and remote configurations. As with Red Hat Enterprise Virtualization and Red Hat Gluster, integrating Red Hat Ceph Storage with Red Hat OpenStack Platform in a hyperconverged model can again deliver lower cost and more streamlined and efficient management in such use cases.
We had a number of sessions on each of these topics at the recently concluded Red Hat Summit. Some of the Red Hat sessions can be found and searched here (registration or login may be required). Check back for information on how you can view recorded sessions and slides. In addition, you can learn more at: redhat.com/storage
About the author
Sean Murphy joined Red Hat in 2014 and is the Senior Principal Product Manager for Data Services & Software Infrastructure.