Around The Block Blog
by Nicholas Sweere, Product Marketing Manager, Dell Compellent — July 12, 2011
The release of vSphere 5 and Site Recovery Manager 5 is one of the largest product launches in VMware’s history. This launch adds additional features, customization ability, and provides an even more robust virtualization offering. In support of the launch, Dell Compellent is updating several of its ever expanding list of integrations and will be certifying them with VMware for the launch of vSphere 5:
Storage Center for vSphere 5
With the release of vSphere 5 and Storage Center 5.5.3, Dell Compellent will be fully supportive of all new storage volume/datastore mappings available through vSphere 5. Principal among these new features will be the ability to support and map up to 64TB VMFS datastores. This will allow virtualization administrators to support more, larger virtual machines (VMs) per volume/datastore.
Site Recovery Manager 5 (SRM 5)
Dell Compellent’s Storage Replication Adapter (SRA) for SRM 5 will be available for download when vSphere 5 launches later this summer. The SRA will support new features within SRM 5, like automated failback from a disaster event and the new work flows for planned migration and downtime. This new version of SRM will provide an even more complete solution for planning, automating, and recovering from disasters or planned downtime. Additionally, Dell Compellent will continue its OEM relationship with VMware for SRM 5, allowing all DR recovery hardware and software support to be centralized under a single roof using the Dell Compellent Copilot support organization.
vStorage APIs for Array Integration (VAAI) - Block Zeroing
Block Zeroing, part of the server offloading APIs known as VAAI, is supported now through our Storage Center releases. Block Zeroing allows servers and Dell Compellent to do what they do best, manage VMs and storage respectively. Block Zeroing reduces CPU load and IO on the server and reduces time to create and manage storage assets for the user.
vStorage APIs for Array Integration (VAAI) – Unmap
The release of vSphere 5 unveils another VAAI offloading API, known as Unmap. Before Unmap, deleting data from VMFS datastores, like a file or VM, did not necessarily allow for that once consumed data to return to the available storage pool. Using Unmap, the server will now notify the Compellent array to free the once consumed storage assets back to the available storage pool. This provides even more storage efficiencies using a Dell Compellent array.
Upcoming Support
We will also be updating the immensely popular and easy to use Dell Compellent vSphere Plug-In, the Enterprise Manager / vSphere Integration, and support for all the vStorage APIs for Array Integration (VAAI) primitives. We are working hard to ensure all our offerings continue to offer customers the quality experience with VMware they expect from Dell Compellent Fluid Data. Check back for updates and announcements on additional VMware integrations. In the meantime, watch Darren Thomas, Vice President and General Manager of Dell Storage, discuss Dell’s alliance with VMware — and how it enables the cloud.
by Scott Horst, Executive Director of Marketing, Dell Storage — October 28, 2010
I was just at the Midsize Enterprise Summit in San Antonio where I did 6 boardroom sessions in 2 days. Each boardroom consisted of about 20 end users that listen to your pitch on your product and ask questions about how things really work. I always prefer to do these sessions in conjunction with a Compellent customer as they can tell the story on what really matters to IT. I did three sessions with Greg Edwards from Harris County, and I did another three sessions with Chuck Matulik from Cross Country Home Services.
I found that customers of all sizes continue to be very frustrated with the rigid boundaries of their existing storage, to the point of needing to consider a new SAN in just 2 years, as their systems don’t scale beyond today’s needs. Compellent knew from the start that scaling beyond is what every business required. As we prepare to introduce the next version of Storage Center, which will encompass the largest hardware release in our company’s history, we are taking the first step in a series of new product releases designed to drive scale and enterprise deeper into our product. With this upcoming release, our scale up story just gets stronger. Customers can continue to add drives, drive types, more spindles, and more powerful controllers, all to scale to meet the demands of their business. We're talking about the ability to deliver more performance, less hardware and lower power in the same footprint using 2.5-inch SAS drives, and eventually as we add infrastructure to our system, doubling drive counts over time. When we deliver this release, we’ll show you all the details.
Our open and agile Fluid Data architecture makes this transition straightforward and seamless – offering backward and forward compatibility. And with this new release, our scale out story becomes bolder. Enterprise Manager serves as that common framework to manage multiple systems as one, and it also provides dozens of points of integration with leading technology. For example, in August at VMworld, we announced our vSphere 4.1 client plug-in. You will be able to manage Compellent storage, snapshots and replication, all through the vSphere client. (Read more about the plug-in capabilities in a recent blog post from John Dias, storage architect for Compellent.)
A key enabling technology for enterprise to scale-out in this new release is Live Volume. Since we introduced automated tiered storage to the market in 2005, Compellent has enabled the automated migration of storage between arrays. In the past, we were intelligently moving data just within the array – now, we're also automatically moving volumes between arrays to provide continuous access and high availability to the business. We will go deeper on this feature later.
So for most companies, all of this would be impressive. But for Compellent, scale up and scale out is not quite enough. Scale Beyond is where we're headed. Over the next year, we intend to drive scale into every dimension of the product – integration, performance, availability, policy, persistence, and more. This is a message I just delivered to our entire channel on our recent “Field First” quarterly webinar. It may appear to some that this increased focus on scale and enterprise is taking us in a new direction, but this has been the plan all along. If you ask the two customers who presented with me at the show, they would tell you that we are unique in the industry in our ability to meet the needs of enterprises of all sizes. Our features were developed to be enterprise from the start, while our platform was built to scale from several terabytes to multiple petabytes. In fact, just to compare, one of the customers I presented with at the show has more than 800 TB’s installed, and the other has less than 100 TB’s – but both are using the same single model we sell to every customer that scales from very small to very large.
The industry alternative to this type of scalability is painful. End users face that real-world cost and hassle of a fork-lift upgrade to a new model, and paying for software all over again. So to Compellent, scale means you can simply upgrade a component at a time, as new technologies arrive, and you can do so without data center disruption and without repurchasing software licenses. That's why technology refreshes with Compellent are actually refreshing, and that’s part of what Scale Beyond is all about. We’re working hard to turn this vision into a reality, and you’ll hear more from us on Scale Beyond soon.
by Tim Plaud, Principal Storage Architect, Dell Compellent — October 12, 2010
My name is Tim Plaud and I'm a principal storage architect with Compellent, a partner of Intel. An evolving set of concepts for the better part of four decades, cloud computing has evolved from being a marketing buzz word with little clarity and even less credibility to a real IT strategy built on a set of proven technologies for data centers. Cloud computing now has a stronger-than-ever foothold and it only seems to be getting stronger.
Many end users I talk to understand the value of cloud services, whether the infrastructure to deliver IT as a service is built on-premises or owned by a service provider. However, the big issue for these companies is how they can go about architecting a solution. To help these companies build their own private clouds, we’re working with Intel and VMware to demonstrate a “Cloud on Wheels” solution in the Intel booth (#206) at SNW, taking place during expo hours on Tuesday, Oct. 12 and Wednesday, Oct. 13. Christian Black, researcher and architect, Intel IT, and I will be in Dallas to drive the demos and talk about how the different technology building blocks go together.
I’m excited to have worked with Intel and VMware on Cloud on Wheels, which will combine Intel Xeon 5600 processors, a converged IP/storage fabric on 10Gb iSCSI, VMware virtual servers and vSphere, and Compellent Fluid Data technology. The demo will illustrate how end users can experience the performance advantages of wide-striping and dynamic RAID, as well as the efficiency advantages of thin provisioning and automated tiered storage. In fact, automated tiered storage plays a particularly important role in this private cloud because application data always finds the right level of storage performance and capacity, without requiring any manual intervention. That’s critical for simplifying administration especially if organizations are tight on resources and don’t have a lot of time to manage storage.
Naturally, virtualized storage resources are just one piece of the private cloud puzzle. Cloud environments require the processing power, the virtual servers and the network connectivity to handle scalability, performance and management requirements. With Intel, VMware and Compellent technologies making up the foundation of the Cloud on Wheels solution, all the pieces work together in a self-tuning data center. Customers can manage large pools of IT resources, including networking, storage and servers, as part of an integrated modular, flexible and scalable cloud requiring very little management. After all, isn't that what this whole cloud thing is all about?