Virtual Iron announced version 4.2 of its server virtualization platform today, adding new features that inch the offering closer to feature parity with market leader VMware ESX Server and that represent firsts for a Xen-based platform. The features include the following:
- multipathing for Ethernet and Fibre Channel networks;
- LiveSnapshot, a space-efficient way to create snapshots and clones of running virtual machines for hot virtual machine backup; and
- storage capacity management features that allow virtual disks to be managed (increased, decreased, added) without downtime.
The new version also adds support for Red Hat Enterprise Linux (RHEL) 5 and Novell SUSE Linux Enterprise Server (SLES) 10. It also includes support for “unmodified” native versions only, shunning the early “paravirtualization” mantra of the Xen community. “Our philosophy has always been around supporting unmodified OSes,” said Mike Grandinetti, Virtual Iron chief marketing officer. “Are you really willing to break the certification of your entire stack for a 1% performance improvement? It would have to be [a] pretty mission-critical [application].”
Virtual Iron also announced that its Virtual Server tools are now packaged as an ISO image, appearing to the administrator as a virtual CD-ROM, for simple deployment. The firm also announced that its platform had been certified for use with storage systems from Network Appliance, which is currently in use at Virtual Iron customer Meganet Communications.
Designing the storage to go along with your virtual environment? In this tip, Anil Desai explains a variety of ways to avoid storage (I/O) bottlenecks. In his analysis, Desai covers all the bases: analyzing a virtual environment’s I/O characteristics, designing RAID configurations and fault tolerance and, finally, planning for host and guest-level backup.
Meanwhile, over at SearchVMware.com, Scott Lowe regales us with a great tip on the ins and outs of VST, EST and VGT VLAN tags in VMware ESX. He explains why VST tags are usually your best choice but also describes cases where EST and VGT might be more appropriate. The Internet is awash in virtualization content, but IMHO, this tip really exemplifies why TechTarget launched SearchVMware.com: to provide VMware administrators with the nitty-gritty technical information they need to properly configure their systems. I hope it’s helpful to you.
Editors’ note: Virtualization Log is a regular feature of the Server Virtualization Blog, where we recap the news, tips and columns that were recently published on SearchServerVirtualization.com and sister TechTarget publications.
Can you get your core hardware vendor to sell you the hardware, the virtualization software, integration services, and support? Sure can. This becomes more of a benefit as you can get direct access to better resources for support and pre-sales matters. One example is to conside Dell as your single stop vendor for your pre-sales, hardware, VMware software, support, and post-installation professional services.
The Dell and VMware Alliance and the VMware from HP programs are good examples of a single-stop shopping resource you can use to streamline your process – and the best part is you have the same account/sales reps as you would for your other hardware and software need s. This could also include your notebooks, laptops, printers, as well as your virtual infrastructure. Anything you can do to make your job easier without sacrificing service and quality is a good idea.
Further, the real benefit is that you can get better pricing than you may direct from the providers – and more direct access to local resources for additional pre-sales support that can enable you to make better forward decisions.
Known for its physical-to-virtual (P2V) migration and capacity planning tools, PlateSpin Ltd. has claimed a virtualization first: a turnkey disaster recovery hardware appliance based on VMware virtualization.
Announced today, PlateSpin Forge consists of a dual-processor Dell server equipped with 2.5 TB of SATA RAID storage, VMware ESX Server virtualization, the underlying PlateSpin PowerConvert P2V software, and the PlateSpin Forge user interface. Together, these components can protect up to 25 “workloads,” running either Windows or Linux, within a VMware Virtual Machine Disk Format or Virtual Hard Disk format or on a physical box.
“To us, a workload is a workload,” said Cadman Chui, PlateSpin vice president of marketing. Forge, Chui explains, simply takes an image of the workload and saves it as a virtual machine (VM). Then, on a scheduled basis, Forge takes the incremental changes that have happened on the VM and applies them. How often Forge takes an image of the workload depends on its criticality.
Recovery, meanwhile, can take place directly on the Forge hardware appliance or back on another virtual or physical host once it’s available.
PlateSpin competitor Vizioncore developed a similar offering this summer called P2V-DR vRanger Pro Module, but unlike Forge, it’s a software-only product, Chui said, requiring customers to buy and configure the underlying hardware.
PlateSpin will sell Forge at a starting price of $50,000, or $2,000 per protected workload. Chui said that the company’s channel partners think Forge will be a big seller because of its ease of use. “You just press the big green button, and away it goes, doing its thing.”
In SearchServerVirtualization.com’s new blog-entry series, “Virtualization book club,” I’ll post information related to our author interviews.
In today’s post, I’d like to introduce you to Eric Hammersley, author of Professional VMware Server: Programmer to Programmer, a previous CIO of the Sassi Institute, and a current engineer at defense and space company EG&G Technical Services in Bloomington, Ind.
Professional VMware Server was the first book published on the VMware product. It offers information on how to install and configure VMware Server; tips for creating base images; methods for organizing image libraries and preparing them for use; best practices for using VmCOM, VmPerl and the application programming interface; and integrate VMware Server in an existing environment and more.
Much to his welcome suprise, Hammersley was contacted by Wiley publishers to write the book.
“I was honored to have the opportunity to write Professional VMware Server,” he said. “Thankfully, it was an easy sell, and I was given wide latitude in terms of direction and content.”
Hammersley first fell in love with virtualization when he worked as a government contractor for the U.S. Navy. He was installing multiple servers and switch racks aboard U.S. Navy ships to support the Navy’s Information Technology for the 21st Century initiative (IT-21).
“With the virtualization technology being developed by VMware and impact on space, the recoverability and survivability of the infrastructure could be greatly increased,” Hammersley said. “I do not know if they ever leveraged virtualization technology as part of that program, but the robustness of the technology today would be a great asset to that ongoing project.”
Although Hammersley doesn’t have other books on the market, we should expect to see his name on the IT bookshelves sometime soon. “We always have Vix to expand on,” he said. “We’ll see what happens.”
Ironically enough, the original VMware Server author missed VMware’s IPO this past summer.
“I never get in on those things in time,” Hammersley joked. “I missed the Nvidia spike back in 2000 that I knew was coming but procrastinated, the Google IPO and now the VMware IPO. I suppose I’ll just be content with writing books and putting in my eight hours.”
With all the hype and hoopla surrounding virtualization, picking the right platform can be a scary proposition. For those of you who haven’t yet chosen VMware’s, Citrix’s or Microsoft ‘s virtualization platform, IT consultant Steve Shah explores their respective virtualization approaches (rather than their technical differences) to help you decide which one is right for you.
If you do go down the VMware road, we have a bunch of tips to get you started. Expert Rick Vanover offers advice on ensuring a successful mass physical-to-virtual migration. The takeaway: Measure twice, cut once.
Meanwhile, for the cheapskates among us, Harley Stagner gives us a two-part tip on how to get failover and high availability for a file server running on ESX without buying VMotion or VMware High Availability. In the first part, Stagner explains how to set up Windows Server 2003 as a file server on an iSCSI storage array; in part two, he describes a nifty way to use the free VMware Converter 3.0.1 and p2vtool.exe to script the failover of that file server, without incurring downtime. Neat-o.
Finally, expert Andrew Kutz helps a VMware user struggling with a script to help him shut down and power up VMs outside VirtualCenter. The answer, Kutz says, lies in the FindByDnsName method. Kutz also explains how to create a library assembly (DLL) that can be referenced from other .NET applications.
When you think virtualization, VMware comes to mind as the leader, correct? Sure ESX is the premium product right now for x86 virtualization, but there is a movement that needs awareness right now. Last week I mentioned that we should evaluate Citrix XenServer and this week I will expand the scope of that recommendation. The base technology of virtualization will soon be a commodity, and basic elements are free with VMware Server, XenServer Express Edition, Microsoft Virtual Server, and Microsoft Virtual PC. The base virtualization technology is now readily available so many ways, that the real distinguishing factor will be the management of the virtual environments, the high-availability, costs, and ease of use.
Once shunned out of many IT shops in lieu of the “Windows Revolution” Novell now offers a virtualization management layer, Novell ZENWorks Orchestrator. Now, before you blow off to some other post – consider this – most Novell products are really good at what they do. NetWare was a superior file server (sure – there were client issues and interoperability issues – but there still is no better rights assignments for file serving), Novell jumped on the Linux boat early, and you can see how Linux has clearly maintained its momentum. So, from the management standpoint we will really need to evaluate this solution as well. Orchestrator also is going to embrace cross-platform (Xen, VMware, and Microsoft) management. That alone should be enought to get your ears perked up. Remember, virtualization is relatively young in the x86 space – so anything we can do to not close any doors from the intial embracing of the technologies would be a good idea.
Just as it said it would back in October, Sun Microsystems Inc. is releasing version 1.0 of its “new” xVM Ops Center management suite, which is really a merger of its Sun N1 System Manager (N1SM) and Sun Connection configuration management software packages that have been brought under the auspices of Sun’s nascent virtualization brand.
Nevertheless, xVM Ops Center Director of Marketing Oren Teich claims there’s more to Sun’s new management bundle than branding. There’s a nice new Web user interface, of course; but perhaps more to the point, xVM Ops Center can manage a far greater number of operating system instances than its prior incarnation. To wit: N1SM maxed out in the 100- to 250-node range; Sun Connection gave up around 500; but Teich claims xVM Ops Center is able to manage around 5,000 running OS instances.
“That’s a lot bigger than any of our customers have calls for using,” Teich said. That includes the Texas Advanced Computing Center, whose 3,936-node Sun-based supercomputer will be managed by xVM Ops Center.
That surge in scalability stems from a redesign of xVM Ops Center’s network architecture. Taking its cues from the Web feed format Really Simple Syndication (RSS) we all use to read frequently updated content like blogs or news, xVM Ops Center shuns the old model of a central management server regularly polling its charges and instead adopts a subscription model where agents take the initiative to report up to the mother ship — the satellite server — via distributed proxies.
“The agents talk to proxies, and the proxies talk to the satellite servers,” Teich said, not the other way around.
Sun’s xVM Ops Center agents also mimic RSS in its use of XML over HTTP. This approach, Teich said, helps Ops Center “avoid having to punch holes all over your firewalls, which has always been a nightmare.”
But as impressive as this 5,000 number may be, you’ll have to wait until the second quarter of 2008 and version 2.0 for Sun xVM to actually be able to manage xVM, the Xen-based hypervisor. “Version 1.0 doesn’t include Xen management; that’s because we don’t have it yet,” Teich said. For now, xVM Ops Center is limited to managing physical machines running Linux or Solaris [Windows support is forthcoming, although no date has been specified]. For tasks within a virtual environment, “customers today would probably use VMware [VirtualCenter].” That’s OK, though, because for now, about 75% of Sun’s “many hundreds of customers” for N1SM and Sun Connection are still running in a physical environment.
Sun has released pricing for xVM Ops Center: $10,000 for the central satellite server and $100 to $350 per managed instance, or guest.
Server consolidation via virtualization can save an organization so much money that people sometimes forget about ways to stretch a virtualization budget further. Not so for Rick Vanover, a tip writer and systems administrator, who offers his thoughts on the most cost-effective servers, network cards, and host bus adapters (HBAs) to buy for a virtualization host.
Speaking of saving money, a local credit union avoided having to spend tens of thousands of dollars on extra disk space for its NetApp storage by running A-SIS data deduplication on its VMware Virtual Disk Infrastructure volumes. A-SIS comes “free” with NetApp NearStore arrays, although of course, NearStore isn’t free.
Got VMs running on SAN storage? In another tip, Rick Vanover details a technique for retrieving a logical unit number (LUN) when the presence of VMware native drivers prevents you from using normal techniques. He also describes the esxcfg-mpath command that you can use from the service console to reveal information about HBAs, multipathing, LUN configurations and the like. Read this; your storage administrator will love you for it.
Today marks an host-oric day, as the first virtual desktops are ready in the lab for my most forward-thinking users (and, as temporary machines, any who happen to suffer hardware failure). As my company is a mid-sized firm, taking this plunge is a bit of “bleeding edge” for us, but it’s too promising to pass up. The early test environment was pretty basic – a few desktops with souped up memory, CentOS 4, VMware Server, and our XP build. First a side-note on CentOS – I love CentOS because it’s almost 100% binary compatible with Red Hat Enterprise Linux. In fact, it’s compiled from their SRPMs, with the copyrighted materials (the logo, some artwork, etc.) removed. On the client-side, ThinStation or any of the other many thin-client linux distros meant to communicate via RDP will work just as well (perhaps better). The roots of this initiative lie in my wanting to have my XP desktop available from where I was — my Macbook Pro, My Freespire 2 desktop, or my Vista desktop. All have desktop virtualization on them, but since they don’t all have the “same” products, mounting a share somewhere wasn’t going to work — and performance might be a bit… underperforming.
The best route was to have it available via RDP. I also wanted to build virtual desktops rfor users. The result, to kill an old commercial’s memory, is that VMware got their peanut butter in my chocolate, or I got my chocolate in VMware’s peanut butter. Either way, I liked the results. It was simple enough to do, and it performed well under even the limted circumstances. Best of all, it’s not complicated to manage. ESX and VirtualCenter more than did the job (though I thinkg a fortune 500 would have need for enhanced management tools, if only for filtering and tracking users to desktops).
After that worked out well for me, I started trimming it back to a more common user-centric desktop build as opposed to the IT-Centric build, taking temporary desktop replacement as a start-in point. The big first was security, while limiting complexity ran a close second. Thanks to AD’s Group Policy handling profile and folder redirection, there’s really no perceivable difference between the user’s original computer and the server-hosted VD. When their PC is fixed, they get it back, and we move on to the next broken-box situation.
The VD solution proved its value there, beating our 2X application server thin-clients (which fared well, but less well than VDs because of the difference in user experience between a linux desktop running a full-screen browser and an XP PC). The next step is to see if we can make this permanent. So, a few IT-savvy first-adopter types are going to get some very old PCs with some very new tricks. I can’t wait…