BACKGROUND IMAGE: iSTOCK/GETTY IMAGES
Large data centers are experiencing massive growth in storage traffic and storage capacity requirements. To enable data center convergence, having software-defined networking and software-defined storage work together is critical to enable automated storage provisioning and the effective management of storage traffic.
Software-defined storage (SDS) provides software-based virtualization, provisioning and management of data storage independent of the underlying hardware. SDS provides software to manage storage deduplication, replication, snapshots, backup and other storage-specific requirements.
The rise of compute virtualization, increased server performance and the popularity of low-cost storage options provide IT with a breadth of options to efficiently deploy storage capacity in today's data centers. Via storage virtualization, SDS offers IT flexibility to allocate storage where and when it is needed. Administrators can upgrade physical storage hardware without disrupting applications access. SDS also simplifies provisioning and management through centralized portals to manage shared pools of storage.
SDS to impact data center convergence
The massive increases in compute capacity (multicore) and complexity of modern applications that include video, Internet of Things (IoT) and big data have driven a huge storage capacity build-out where a petabyte scale is now normal. Flash storage, hyper-converged systems and protocols including virtual storage area network and Hadoop place tremendous requirements on the data center network. Storage traffic is now moving east-west (as well as north-south) and often has the requirement for low-latency access to data.
Data center networks must offer elastic capacity to handle the growth in storage traffic, rapid provisioning, centralized control and management, security and the ability to support any type of storage with any type of compute and application. Data center convergence means the network must support the coexistence of data (Ethernet) and storage traffic. The network must support a separate, virtual channel for storage traffic to handle bursty traffic and required low latency.
Storage on IP networks introduces several new types of traffic, some creating large volumes of unpredictable traffic into the data center.
SDN helps segment storage traffic
SDN provides a number of key capabilities to help IT organizations build data centers to handle the emerging storage network challenges. SDN provides network virtualization to easily segment data and storage traffic. It offers rapid provisioning of new network capacity and centralized management of network traffic.
SDN can provide visibility into traffic flows, network bottlenecks and provide appropriate capacity to bandwidth or latency sensitive applications. With its segmentation and visibility capabilities, SDN can help secure network traffic by identifying security breaches and assisting in remediation efforts. For greenfield networks or network upgrades, SDN offers the option to deploy cost-effective white-box switches to upgrade bandwidth to 10 GbE, 40 GbE or more.
SDS suppliers enabling data center convergence
IT professionals have a wide range of SDS solutions to implement storage virtualization and SDN products to enhance data center networks. These include IT, storage, hyper-convergence and network suppliers large and small. A selected list of these suppliers includes:
- Hewlett Packard Enterprise
- Red Hat
- Pluribus Networks
- Big Switch Networks Inc.
- Jeda Networks
- Coho Data
- FalconStor Software
SDS recommendations for IT leaders
Private cloud infrastructures and applications including video, big data and IoT require new approaches to data storage and networking. SDS provides significant advantages for offering a pooled-storage infrastructure that is automatically provisioned to application requirements. SDS also enables IT to deploy cost-effective commodity storage managed by software running on commercial servers, thus providing a rapid scale-out of storage capacity.
Converged, private cloud architectures will employ converged, virtualized networks to segment, provision and manage storage as well as data traffic. SDN provides the networking capabilities (provisioning, management and visibility) to enable virtualized storage.
A key challenge for IT leaders is sorting through a plethora of SDN and SDS options to build their next-generation converged data centers. Both SDN and SDS are in the early stages of data center deployment. And despite ongoing efforts, there are no clear blueprints for deploying comprehensive software-defined data centers.
Using orchestration tools for software-defined storage and SDN
Get the low-down on software-defined storage
Hyper-converged infrastructure makes storage cool again