Storage Where You Need It, When You Need It

Summary:According to industry analysts, the amount of data being directly managed in enterprise datacenters will grow 14 fold over the next eight years. That’s a lot of information. It also implies a lot of infrastructure. But there is an irony here.

According to industry analysts, the amount of data being directly managed in enterprise datacenters will grow 14 fold over the next eight years. That’s a lot of information. It also implies a lot of infrastructure. But there is an irony here.

As most CIOs understand, the business value of the IT department is only loosely correlated to the infrastructure they manage. The real value is the information contained in the datacenter. And that data is not worth much if the data is not safe and accessible. Therefore the datacenter needs to be architected in a way that stored information is highly available and applications consuming it have efficient and reliable access. How do you do that?

One way is to have massive numbers of data storage devices directly connected to each other. This creates problems related to managing so many appliances and inefficiencies in the way they are used. It also does not really address some fundamental operational issues. In a dynamic and complex environment, the flexibility to move workloads between servers in the datacenter is only a partial solution. There is also a frequent requirement to reallocate storage between pools without disrupting services. This may be necessary to improve latency, re-balance storage devices, or undertake maintenance.

Another way to address management needs is to pool storage from multiple devices into what appears to be one virtual device. This is storage virtualization.

Storage virtualization can help achieve greater allocation flexibility and higher levels of reliability by acting as an abstraction layer that redirects read and write requests from virtual to physical storage devices. Because applications only interact with a logical representation of the devices, they are able to continue functioning seamlessly—even if data is transferred from one physical device to another.

The most common technology that implements storage virtualization, particularly in large enterprises, is a storage area network (SAN), a dedicated network providing block-level storage usually running over the Fiber Channel or iSCSI protocol. The virtualization functions can either be performed by the SAN appliance itself or the network switch that interconnects the SANs and the hosts.

SANs are an excellent option for corporations with stringent needs for performance and reliability. However, they are also expensive and complex to configure. Windows Server 2012 has a new capability called Storage Spaces, which provides a storage solution using commodity hardware.

The technology works at two levels. Storage pools aggregate physical disk units into virtualized administration units. There is complete flexibility in the combination of disks, which can employ different technologies (e.g. SATA or SAS), operate at different speeds, and provide different capacity, but are combined to form a single homogenous pool. Within these pools, it is possible to allocate virtual hard drive images, called Storage Spaces. In addition to the desired capacity (which may be thinly provisioned), the configuration can specify a desired level of resiliency and administrative control.

Pools allow some degree of reshuffling within the pools, but sometimes it is necessary to transfer data to other targets. This transfer of data should occur without disrupting the services that are using them. A technology that minimizes the impact on the host computer is Offloaded Data Transfer (ODX). ODX is a technique that enables intelligent storage arrays to transfer data between compatible storage devices directly, without transiting the server. It makes it possible to move large chunks of data, such as virtual machines, at very high speeds with minimal latency, while freeing up the server’s CPU and network interface for other operations. Note that with Cloud-integrated Storage (CiS), these functions become totally automated across a hybrid cloud.

Moveable storage is essential for high performance, mission-critical applications that cannot afford any downtime. It enables seamless operation and ensures that server performance does not degrade while data is on the move. By selecting the right storage technologies, it becomes possible to manage more effective use of higher volumes of data.

Topics: Data Centers


John Rhoton is a contributor to CBS Interactive's custom content group, which powers this Microsoft sponsored blog. He is a technology strategist who specializes in consulting to global enterprise customers with a focus on cloud computing.His tenure in the IT industry spans over twenty-five years at major technology companies, defining an... Full Bio

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Related Stories

The best of ZDNet, delivered

You have been successfully signed up. To sign up for more newsletters or to manage your account, visit the Newsletter Subscription Center.
Subscription failed.