It's familiar as reinventing the wheel: Companies head down the path to virtualization starting with static or dedicated servers, then do OS virtualization for test and development, followed by some form of data center consolidation. At each step along the way, they are left to deal with the same old problem--server management, which only gets worse as they delve deeper into virtualization.
Now, says Ken Salchow, manager of Technical Marketing for F5 Networks, customers are looking to break out of that old mold via the virtues of on-demand services--just click a button and there's your app, which puts an end to the inflexible mapping of users to physical machines.
"Customers want the control and capabilities from their traditional infrastructures, along with the simplicity and flexibility to outsource and make their applications totally virtual," Salchow declares, adding F5's solution to this dilemma is the dynamic services model, based on the company's newly introduced BIG-IP Version 10.2 software that helps extend enterprise data center architectures to the cloud, and enables on demand services.
F5 takes a holistic approach, which integrates multi-vendor solutions into a unified whole--and raises the specter of finger pointing in the cloud. Salchow, however, says not to worry, because no one vendor can do it all. As he puts it, "We can't solve all these problems alone, it's an ecosystem effort, and it's a matter of trying to tie all these things together. There's a lot of fear about what the cloud means."
This translates into F5's use of a virtual Application Delivery Controller (ADC) platform which "in combination with physical ADCs gives customers the unique ability to seamlessly extend their current operational model into multiple data centers, hosting providers, and cloud environments on demand, while utilizing the same trusted configuration, features, and control that they have developed over the years of application delivery deployment."
F5's new BIG IP Edge Gateway fits in by making it possible for enterprise users to simplify managing app services such as access, security and optimization within the company's new dynamic services model.
Salchow admits that there are still serious challenges to overcome, such as slower data rates between the cloud and the data center than users are accustomed to with their data center links. This means it takes longer to move images, which can become very large and unwieldy. He says that the smallest VM he ever built was 7 gigabits, "and that was just a workstation. You just can't move images fast enough." F5 aims to alleviate that pain by providing dynamic optimized tunnels that eliminate the manual processes involved in moving VMs from one location to another.
"Manual processes just don't cut it," he comments.
Posted by Bruce Hoard on 04/22/2010 at 12:48 PM2 comments
F5 Networks is a pretty cool company. Anybody who offers a solution that maintains a user connection to a VM moved half way around the world would at least seem to have the mobility thing working for them, which puts them on top of the management stack, so to speak. I offer these 10 questions from their PowerPoint playbook designed to help you determine if your cloud provider is enterprise-ready.
- How do they instrument and report to ensure end-user performance SLAs are met?
- How do they enable access control and ensure it remains with the enterprise?
- How do they enable network and application security, and demonstrate that enforcement?
- What models do they provide for disaster recovery, elasticity, scale and high availability?
- Can the enterprise easily mimic their existing architectures, still maintain control and easily move between cloud partners?
- What applications and types are supported?
- Are systems set up to simplify, secure and optimize virtual machine transfers between the enterprise and the cloud?
- How do they automate or orchestrate management between the enterprise and cloud provider?
- What attributes regarding the end user and application (device, location, role, performance, availability) are incorporated into making intelligent traffic management decisions?
- For all of the questions above, if a provider has a solution, are they different or deficient compared to what the enterprise is already doing? Is the enterprise required to support a different model for each provider?
I'll be blogging more on their latest "vision announcement" tomorrow. Stay tuned.
Posted by Bruce Hoard on 04/21/2010 at 12:48 PM6 comments
Despite the economic travails of the country, times are good for i/o Data Centers, which has more than 300 customers, is rapidly selling capacity at its 125,000 square-foot facility in Scottsdale, Ariz., and has built out 180,000 square feet of raised floor at its 538,000 square-foot facility in Phoenix, which opened last June, and is the largest commercially available data center in the U.S.
"We operate on an inverse path from the economy," says Jason Fererra, director of marketing. "We've really had a great few years."
i/o Data Centers, which has been in business since 2007, guarantees 100 percent uptime and is carrier-neutral. As a result, 12 telecom providers are currently linked into the company's two data centers, which are 30 miles apart and connected by a fiber ring. The company offers 20 megawatts of conditioned power, which it says is "clean power that i/o delivers to its collocation and outsourced data center customers."
In addition to its data center and collocation offerings, i/o Data Centers recently introduced an on-demand service that provides the ability to access enterprise-grade infrastructure processing, including virtualization, storage and bandwidth. "We allow our customers to take what they need when they need it," Ferrera notes.
Highlights of the on-demand service include:
- Access to VMware's vCenter controls
- Best-in-Class servers
- Unlimited SAN capacity
- Up to 32 GB RAM per server
Although i/o Data Centers has competition in the form of companies like Equinix, Digital Realty Trust and Dupont Fabros, Fererra says their competing data centers are more distributed, while i/o Data Centers prefers to concentrate its resources in larger, centralized facilities.
Question: Do you feel comfortable outsourcing your company's data?
Posted by Bruce Hoard on 04/20/2010 at 12:48 PM6 comments
Now that desktops live in the data center, things have changed dramatically for admins and users alike. In order to ease everybody into the new world of VDI, Citrix CTO Simon Crosby urges companies to roll out pilots before taking the plunge into full-blown, enterprise-wide implementations. One good way of doing that, in his estimation, is by delivering virtual desktops from the cloud.
"In general, acquiring a reasonably large number of servers to deliver desktops to your user base imposes additional resource requirements -- space, cooling, and capital outlays for the enterprise," Crosby declares, adding "that may be beyond what they want to do, so the notion of hosting virtual desktops in the cloud is very attractive."
This way, he suggests, it is possible to develop a cost-efficient proof-of-concept by simply going to a cloud provider, bringing up your data, renting a VM by the hour and turning it off at night after the users go home. "So I personally think that virtual desktop infrastructure will have a strong element in it which is related to the cloud," he says.
Even though he admits there are some "challenges" involved in this approach, Crosby assures us that Citrix is working hard to get cloud computing vendors equipped to deliver desktops as part of the "broader go-to-market."
For companies just getting started in desktop virtualization, Citrix is offering an Express Edition of XenDesktop 4 that includes 10 free licenses. It’s available here.
Question: Do you think it is widely feasible to provision VDI from the cloud at this point?
Posted by Bruce Hoard on 04/14/2010 at 12:48 PM1 comments
Building on its virtualization systems management foundation, Reflex Systems introduced vProfile, a configuration management tool that enables users to visually isolate configuration problems while insuring that internal and external configuration standards meet compliance requirements.
As part of the vendor's flagship Reflex Virtualization Management Center (VMC) family, vProfile competes with the steadily increasing number of graphics-oriented virtualization management packages entering the market by offering what it calls "heat maps" that depict geographical disparities in virtual infrastructure environments.
In addition to vProfile, other VMC modules include vTrust and vWatch Monitoring. All these modules are connected via a common VMC management console that Reflex Systems says "abstracts and stores information about the virtual infrastructure in a virtual CMDB and presents it in an easy-to-use graphical interface."
"It is a huge value that these modules are not point solutions but are all integrated," says Mike Wronski, VP of Product Management.
Reflex VMC reportedly tracks thousands of properties associated with hosts, VMs and virtual management servers such as VMware vCenter. vProfile uses storage, networking and virtual management servers and operational configuration types for hosts running hypervisors such as vSphere and ESXi. According to Wronski, the company is also considering adding compatibility management with Hyper-V and Xen hypervisors, although he says there is not currently sufficient demand to do so.
At first glance, VMC's sophisticated graphics seem similar to those offered by Xangati, another management vendor which recently debuted its latest virtualization management package for vCenter environments. Asked to compare the two, Wronski said that Xangati stresses performance monitoring while VMC is more oriented toward operations and configuration management.
Reflex says vProfile configuration management plots VM configuration changes according to a "heatmap interface that uses a color schema to signify disparities--the more configuration differences there are, the 'hotter' a particular area of the map shows the differences. In addition, users can leverage customized pivot tables to cross-tabulate and summarize configuration information to matrix different combinations of properties. Wronski says "There is a huge array of metrics to choose from."
He goes on to state that while Reflex and VMware have been development partners since 2006, the two companies are in "coopetition," especially with VMware's Host Profiles product. "We still work closely with VMware," he says, "and there are some gaps they leave open for us."
Pricing for each module, including VMC platform functionality, is $795 per socket, which the company says is "very similar" to the VMware pricing model.
Posted by Bruce Hoard on 04/06/2010 at 12:48 PM1 comments
Now that just about everybody agrees that hypervisors are becoming commoditized, and management is the money-making mantra, the number of vendors delving into this lucrative realm is riding a high growth curve.
You'd think that with a name like VKernel, a company would be targeting virtualization players like Red Hat and Citrix, but the firms being targeted by this vendor are VMware and Microsoft (Hyper-V).
VKernel, which provides capacity management tools for virtualized data centers, is introducing Capacity Analyzer 4.3, which it claims is the first capacity management package for VMware and Hyper-V. Capacity Analyzer 4.3 enables users to manage both VMware and Hyper-V environments from a single console, which is pretty much de riguer for any product in this class.
More specifically, the product enables Microsoft Systems Center Operations Manager and Virtual Machine Manager users to receive alerts on performance issues in their VMware vCenter and ESX environments.
According to VKernel, Capacity Analyzer "automatically collects performance and capacity data , profiles each virtualized application and builds an accurate model of available capacity across servers and storage at every layer of the user's virtualized infrastructure. Capacity Analyzer's extensive analyses then identify capacity bottlenecks that are impacting performance, predict where future problems will arrive, give specific instructions on how to resolve these issues and tells users where to place new virtual machines to avoid capacity issues."
Capacity Analyzer 4.3 costs $299 per socket.
Posted by Bruce Hoard on 04/01/2010 at 12:48 PM2 comments
When RingCube's vDesk 2.0 came out last September, our reviewer Chris Wolf of the Burton Group called it a "massive improvement" over 1.0. He lauded the workspace virtualization product for its Active Directory integration, its ability to join vDesks to a domain, and its "straightforward, policy-based management" capabilities.
Eager to move to the next level, RingCube has now unveiled vDesk 3.0, which the company says "simplifies the creation, access and management of Windows desktops" by introducing management functionality that delivers global scalability, adds faster MobileSync to the equation and provides Windows 7 support. The company also says that as a result of its collaboration with Intel, it is offering hardware-assisted isolation control with Intel vPro technology.
RingCube is taking on the Citrix XenDesktops and VMware Views of the desktop virtualization world via its ability to deliver virtual workspaces to large enterprises, delegate IT management and efficiently distribute storage. In order to nail down that claim, the company quotes Mike Prepelica, director of information technology, Revere Electric Supply Company, who praises vDesk 3.0 by saying that after he looked at solutions from "large VDI vendors" he favored the product's simple, cost-efficient deployment of vDesk to remote employees.
According to Doug Dooley, RingCube VP of marketing and product development, "We virtualize on top of Windows, there's no second OS and it runs extremely fast--99.8 percent of what the host can deliver. It's centrally managed and stored, and locally executed." He goes on to say that vDesk 3.0 employs a hybrid approach, wherein virtual desktops are stored on file servers and streamed across the network to local endpoints.
The hardware-assisted isolation control in 3.0 provides increased workspace security when running on PCs with vPro virtualization technology. Of the collaboration with Intel, Dooley declares, "Intel gave us full access to their engineering department."
vDesk 3.0 is available immediately from RingCube, and pricing starts at $200 USD per user for the Standard edition and $300 per user for the Enterprise edition.
Posted by Bruce Hoard on 03/31/2010 at 12:48 PM2 comments
While faxes may seem passé to millions of digital users, fax technology lives on at corporations around the world, where fax servers continue to play valuable roles.
According to FaxBack, Inc., every day, users fax mission-critical business information, including contracts, legal documents, sales quotes, purchase orders, order confirmations, event announcements, membership applications and
much more. For Fortune 500 companies fax is a core communications tool. It is estimated by Peter Davidson, IDC Fax analyst, that over 85 percent of SOHO's (Small Business/Home Office) rely on fax.
Given its seemingly global popularity, it should come as no surprise that fax servers are being virtualized similarly to so many of their corporate counterparts. Enter Biscom, which maintains that virtualizing fax servers leads to better fax server resource management, while making it possible to run workstations and servers with different OSes concurrently on the same fax server or servers. The result: Virtual servers make it possible for users to comply with the mantra of doing more with less.
Here's the Biscom elevator pitch: "All Biscom fax server solutions are designed to achieve a common goal: streamline business processes and improve the bottom line. Whether the need is to fax-enable e-mail, integrate fax with a VoIP network, automate unattended high-volume fax deliveries, or implement fax processing workflows, Biscom works closely with its customers to ensure a quick ROI and savings of both time and money."
For more in-depth information, check out The Essential Guide to Fax Server Software, which is published by Windows ITPro.
Posted by Bruce Hoard on 03/30/2010 at 12:48 PM3 comments
Not all that long ago, virtual labs were something less than sexy. They were largely the domains of development, QA and operations, and without the involvement of higher-level LOBs such as sales, customer support, marketing and technical publications, they cruised along under the radar.
No more, says analyst firm voke, inc., which has released its second study on the topic in the form of a "Market Snapshot Report on Virtual Lab Management (VLM)," which assesses VLM technology traits as ROI, market readiness, use and awareness. It also covers the increasing demand and utilization of the technology.
The study was done between August 2009 and February 2010, and was based on input from an independent survey of 100 organizations using VLM technology.
According to Theresa Lanowitz, lead analyst for voke, inc., "Virtual lab management continues to prove itself as a rare, real, breakthrough use of technology revolutionizing the computing industry and the enterprise. Virtualization remains the defining computing technology of the 21st century."
The study includes both key findings and ROI results. Among the key findings, 100 percent of participants who evaluated commercial solutions ultimately adopted them, and VLM is viewed as a way to change the dynamics of IT organizations. On the ROI side, "48 percent of participants reported direct head count savings and productivity improvements," and "On average, 57 servers per organization were eliminated."
Question: Would your organization benefit from a stronger focus on virtual lab management?
Posted by Bruce Hoard on 03/29/2010 at 12:48 PM4 comments
SunGard Availability Service provides disaster recovery services, managed IT services, information availability consulting services and business continuity management software to more than 10,000 customers in North America and Europe. As the offspring of SunGard, they maintain five million square feet of datacenter and operations space, and help IT organizations across a wide spectrum of industries deal with their DR needs.
We're talking about a bedrock company here, so when it announces an expansion of its enterprise-grade cloud computing platform to deliver application availability "and ultimately recovery in the cloud," prospective cloud customers would be wise to take note.
According to Satish Hemachandran, director, product management, SunGard Availability Services, his cloud customers have a lot to gain from going with this enhanced enterprise-grade cloud offering, not the least being access to the VBlock Infrastructure Package, born of the Virtual Computing Environment coalition composed of Cisco, EMC and VMware.
"We have a very close partnership with these vendors," Hemachandran says.
According to the company's press release, the Vblock Infrastructure Package is a "preconfigured, pretested solution that combines best-in-class technology, including Cisco's Unified Computing System (UCS), Nexus 100v and Multilayer Director Switches (MDS), EMC CLARiiON storage (secured by RSA), and VMware's vSphere platform. This technology choice...will provide customers with a path to pervasive virtualization and cloud IT infrastructure, while helping reduce total cost of ownership and meeting today's most demanding use cases."
That's a lot of powerful technology, and SunGard Availability Services has a lot of experience working on that level. As Hemachandran puts it, "We basically see enterprise cloud computing as a logical extension of what we've been doing.
Posted by Bruce Hoard on 03/25/2010 at 12:48 PM5 comments
SureBackup from Veeam Software is not just another product--in fact, it's not a product at all, it's a set of features to help the reliability of your backups, and it is embedded in version 5.0 of Veeam Backup & Replication, which is slated to debut this summer.
Here's another non-traditional thing about SureBackup, according to Doug Hazelman, Veeam's director, Global Systems Engineering Group: "Customers don't know they want it." And why is that, we inquired? "They didn't think it was possible, but virtualization has pioneered the capability of image-level backups," he replies. It is, Hazelman adds with a conspiratorial pause, "a dirty little secret."
Until now, that is.
As he goes on to explain, even though virtualization and the advent of image-level backups have brought many changes, one problem remains: Just because a backup job completes successfully does not mean the machine can be recovered and booted. Even if the integrity of the backup file is verified, there is no assurance that the OS and apps will start without errors, or that data will be intact.
The problem is, Veeam maintains, there are many possible issues, both procedural and configuration-related, that can interfere. The only way to be completely certain is to test and verify the recoverability of every backup, which, until now, was too expensive and time-consuming to be feasible.
Enter Veeam SureBackup, which eliminates these long-standing barriers while bringing certainty to image-level backups.
Not only has Veeam found the seemingly unfindable formula for sure backups, but it has also found one that is said to make it possible to verify the recoverability of every backup in a matter of minutes. As a result, says Veeam, while organizations are able to "embrace image-level backups to improve recovery time and recovery objectives, they are also able to comply more fully with 'reasonable measures,' as required by internal and external regulations, such as HIPAA and SOX."
The secret to this solution is technology that allows VMs to run directly from compressed backups. By publishing the content of backup files directly to WMware ESX hosts, Veeam says, the need to extract backup files and the time and storage consumed by them, is eliminated.
It all sounds good, and Veeam got some pretty good analysts to provide supportive quotes, so it would seem that the only remaining challenge is getting the word out ASAP.
As Hazelman puts it, "This is going to take some time for the market to absorb because it is so different."
Question: Do you think Veeam is on to something here?
Posted by Bruce Hoard on 03/24/2010 at 12:48 PM2 comments
Saying "Seeing is understanding," Xangati recently introduced virtual appliances designed to shine a light on all communications across both virtual and physical worlds. These products are designed to make it easier for IT organizations to move from the physical to the virtual world without losing the ability to monitor traffic across both environments.
Xangati for ESX and the Xangati Management Dashboard were specifically created to eliminate "blind spots" that have plagued hybrid IT shops that must maintain a mix of physical and virtual technologies. By viewing virtual communications within a physical hypervisor, users will be able to see problems coming and eliminate them before they can do any damage.
According to Xangati, users can now track 5,000 "identities" in virtual or physical eco-systems, adding, "We can provide a lot of visibility at a price point that virtualization teams have a budget for." The product is specifically designed to complement VMware's vCenter by adding what Xangati calls "360-degree communication visibility into all of the critical elements that it manages."
Xangati says the results of a pilot program it conducted prove that virtual infrastructures are not the cause of "emerging performance issues," and that better physical-to-virtual communications "foster better relationships between virtualization teams and the owners of the servers that they have created."
The new suite of virtual appliances is available now. Xangati for ESX can be downloaded here for a free 14-day trial, or at an immediate cost of $299. The Xangati Management Dashboard Standard edition is available for $4,999, and the Enterprise edition costs $9,999. The company is also offering a Starter Kit available at the discounted rate of $9,999 -- including Xangati for ESX for up to 20 hosts and the Xangati Management Dashboard for Enterprise edition.
Posted by Bruce Hoard on 03/23/2010 at 12:48 PM3 comments
According to IBM, which announced plans to go online with its commercial cloud service for software development and testing, its new open cloud environment also includes support for SUSE Linux Enterprise from Novell.
Smart Business Development and Test on the IBM Cloud will be available in the second quarter of 2010 in the U.S. and Canada, and it will roll out globally throughout 2010.
Posted by Bruce Hoard on 03/23/2010 at 12:48 PM3 comments
Red Hat knows the drill: Citrix commands most of the commercial, Xen-based buzz while Red Hat is viewed by many observers as a marginal, me-too player.
Not so much now that Red Hat announced that it has struck a deal with IBM to supply Big Blue with the open source, KVM virtualization technology that will power IBM's new cloud. Red Hat brass was reportedly crowing about the deal, which they claim is their third big-time virtualization coup of the still-young year, and one which puts them in the thick of the rapidly emerging cloud market.
Commenting to InternetNews.com, Scott Crenshaw, VP and GM in charge of Red Hat's cloud business unit, declared, "IBM could have chosen anything -- they could have gone with VMware, Microsoft's Hyper-V, or something from kernel.org, but instead they chose Red Hat Enterprise Virtualization. We think the evidence is clear that KVM is the virtualization technology of the future, and that RHEV is our way of delivering that with an enterprise-class product."
Posted by Bruce Hoard on 03/23/2010 at 12:48 PM7 comments
Microsoft's announcement today describing new agreements with Citrix designed to bolster their 20-year relationship at the expense of VMware was a huge win for Citrix, which is finally freed from all the unsubstantiated speculation and irresponsible claims that it will somehow be forsaken and devoured by Microsoft.
When you closely evaluate the situation, it simply doesn't make sense for Microsoft to back out of a relationship with Citrix. Hey, don't take my word for it -- ask Brad Anderson, a Microsoft corporate vice president in charge of the company's Management Service Division, who told me, "When I think about the best way to work for us, I use Citrix as the example."
A healthy Microsoft-Citrix relationship only nurtures Microsoft's key markets by providing them with jointly developed solutions that customers can depend on. At the same time, Citrix feeds unabated at the profits trough as it sells its wares to customer after happy customer with Microsoft's blessing.
From the Citrix perspective, what's almost as good as having Microsoft's blessing is VMware's not only not having it, but having to endure insult to injury while Microsoft and Citrix launch two promotional schemes, "Rescue for VMware VDI" and "VDI Kickstart." The former has the effrontery to suggest that disaffected Microsoft customers running VMware View are looking to bail out of their failing implementations, while the latter provides a safe landing pad for said Benedict Arnolds by offering them up to 500 XenDesktop VDI Edition (annual) device licenses, and up to 500 VDI Standard Suite device licenses in exchange for their tawdry View licenses.
Ouch.
What's also sweet for Citrix is the fact that it is bringing a solid contribution to the table with its HDX technology, which has given XenDesktop a leg up on View, and is now being jointly developed by the two companies to "leverage and enhance" Microsoft RemoteFX. Microsoft defines RemoteFX as "a set of RDP technologies that are being added to Windows Server 2008 R2, SP1 that will allow users to work remotely in a Windows Aero desktop environment, watch full-motion video, enjoy Silverlight animations and run 3-D apps -- all with the fidelity of a local-like performance when connecting with the LAN." This is clearly a technology with legs that adds value to Microsoft's desktop virtualization efforts.
And before we get too far away from the topic of not making sense, it's time for everybody to agree that XenServer, which has been flogged repeatedly for not making a bigger dent in the server virtualization market, is not going anywhere, least of all, down the toilet. Why would Citrix drop XenServer, which is: 1) currently being refreshed with a new version, and 2) the virtualization engine behind its past and upcoming stars, XenApp, and XenDesktop, respectively? (Add Netscaler to the upcoming bracket). XenServer might not be a star, but it has a future.
The battle for the virtualized desktop may only be starting, but Microsoft and Citrix have already shown they are in fighting trim.
Posted by Bruce Hoard on 03/18/2010 at 12:48 PM1 comments