Most physical-to-virtual migration tools are about either assessing a migration, assisting with it, or both.
By submitting your email address, you agree to receive emails regarding relevant topic offers from TechTarget and its partners. You can withdraw your consent at any time. Contact TechTarget at 275 Grove Street, Newton, MA.
Assessment involves using an app to examine a physical machine and its software workload, and to make recommendations about the constitution of the target VM. Assisting includes automating the actual migration process, and perhaps also some follow-up after the fact to make sure the migrated machine performs as expected.
When I came across 5nine P2V Planner, I thought it was a tool mainly designed for the former of these two jobs. What’s most interesting about it is how it’s actually a VM-product decision-making tool. When used, it provides an admin with information about the relative benefits of migrating the job in question to either Microsoft Hyper-V or VMware.
What Planner’s really for
P2V Planner is in some ways a younger brother to the 5Nine Migrator product. The latter performs analysis and planning, aids in the migration with P2V conversion, and also provides ongoing after-the-fact assessment of the migrated machines and workloads. Planner, on the other hand, is entirely about the planning and assessment stage, but anyone who’s used Migrator (or intends to) ought to find Planner familiar, as it uses many of the same metaphors and reporting methods.
Planner’s big reason for existence is as a decision-making tool, for those who are assessing whether VMware or Microsoft Hyper-V is the better fit for their virtualization needs. Many admins are probably tempted to say “Just use VMware, it’s what everyone else uses,” or “If you’re already a Microsoft shop, Hyper-V’s the obvious pick.”
The truth may be less cut-and-dried than that. A workload which is highly optimized in one VM may be problematic in another. The costs of licensing may vary drastically depending on what you’re doing. And last but not least, any existing investment you have in a given VM infrastructure is going to come into play as well. To that end, it helps to have some idea what migrating to one solution versus the other will entail in multiple aspects.
Planner uses a number of different calculations to determine the virtualization ROI and TCO for migrating to each option, many of which I’ve touched on above. Memory usage, for instance: you may be able to consolidate more VMs on a given host if the analyzed average memory usage on each host is markedly less than the installed physical memory in the system.
What the other products do
Other migration-assessment and -assistance products, especially those provided by the makers of VM solutions, are designed to do two basic things: 1) assess migration to a single VM solution and b) assist in the actual performance of that migration.
Because Planner puts more of an emphasis on decision-making, it’s not a direct one-for-one substitute for the other migration-planning tools. They may be designed to migrate more than one type of source system -- Windows, Linux, Solaris, etc. -- but the target environment is typically whatever the VM product is from that company.
There are a few exceptions. One of the biggest is Novell’s PlateSpin Migrate, which is designed to migrate to multiple virtualization target platforms. It supports Novell’s own Xen on SLES 10, as you can imagine, but also all the other major (and some minor) players: VMware vCenter, ESX and ESXi; Solaris 10 (u5 and U6) zone servers; Citrix XenServer 5.5; and Microsoft Hyper-V 2008. Some of that heterogeneity is almost certainly due to PlateSpin being a former third-party product which Novell acquired.
That said, PlateSpin doesn’t concentrate as much on the assessment end -- in other words, it isn’t designed to help you figure out which of these target platforms would be the best choice for your workload. Its main focus is the nuts-and-bolts of migration itself, and it assumes you know which target you want and why.
More on P2V tools
Selecting the right conversion tool for your P2V migration
Top five tips for P2V conversion success
Another, more direct product to compare with would be Lanamark Suite, which is designed to help an enterprise put together both desktop and datacenter solutions using both virtual and physical hardware. It’s mainly intended for sales-and-services use, but it competes with Planner in the sense that it analyzes workloads and makes recommendations for specific virtualization platforms. Lanamark also offers a broader selection of target platforms: Citrix, Microsoft and VMware, rather than just Microsoft and VMware alone.
Pride and prejudice
From my own conversations with admins who oversee VM migrations (as well as the day-to-day care and feeding of VMs), I've found that they tend to be fiercely loyal to a given VM product. If they started with VMware, they stick with it; if they’re accustomed to Hyper-V, it’ll be that for them. Most of this is, I suspect, more about personal habit, departmental politics or platform prejudices than the objective technical merits of one platform versus another. Some of this is justified. For instance, I suspect any admin who is concerned with absolute compatibility would feel more comfortable using the P2V tools provided by his particular VM vendor.
That said, If I presented someone with hard evidence that VMware was the better-performing environment for a given workload (or, conversely, that Hyper-V was better suited to the job), they might well resist. After all, if they’re more comfortable with a given platform, wouldn’t that by itself be a boon since it would mean less need to retrain, modify work habits and so on? True, but I’d argue a better-performing VM would be worth whatever extra effort was needed for its upkeep.
So, tools like Planner, which allows an admin to educate himself about what the options are and their relative benefits. I’d like to think the more we have tools like this -- and the easier they are to obtain and use -- the less reflexive prejudice there will be about the use of one VM product over another, and the more such decisions will be driven by data. Of course, I can’t ignore how each product may well have its own assumptions and biases, but that’s what competition is for.
ABOUT THE AUTHOR
Serdar Yegulalp has been writing about computers and IT for more than 15 years for a variety of publications, including InformationWeek and Windows Magazine.