What’s your vMotion Story?
If you have worked in IT for long enough, then you will have a vCenter vMotion story. The very first time you saw vMotion work. And the slow realisation about how it was going to change our jobs forever.
For me, I had been playing around with VMware Workstation and Microsoft Virtual PC, but we were trying out VMware ESX (version 3 I think) in our lab, and my colleague asked if I wanted to see something ‘interesting’.
We had a window open running a continuous ping and the virtual machine moved from one physical server to another with no disruption of service. For one of the pings, a little latency, but the virtual server basically didn’t really blink. I was so flabbergasted, I ran the vMotion for myself. About ten times. At one point, I ran a remote desktop session to the server and kicked off the vMotion while I was typing in notepad. Started switching off hosts to prove the virtual machines had indeed moved and this wasn’t a trick. I was like a kid with a new toy.
vCenter vMotion: The past
Virtualisation only really took off when the vMotion was developed. Being able to migrate a running virtual machine from one host to another with ZERO downtime was what led a lot of people to even consider adopting virtualisation in production and was the turning point for VMware, without vMotion, they would not be the company they are today.
For such a key piece of virtualisation history, there is surprisingly very little information about how it was developed available online. I found this great blog, with a brief story of how it came about and a number of prominent virtualisation bloggers gave their own vMotion stories. So what’s yours?
vCenter vMotion: The present
Things have moved on quite a bit since then.
In a previous blog post, I mentioned that VMware released vSphere 6.0 earlier this year.
As part of that release, vMotion was enhanced to include a new set of migrations:
- Across vCenters
- Across virtual switches
- Across long distances
- With L2 adjacency not being a requirement between source and target vMotion networks
This added functionality takes a great feature and makes it even better. By removing the boundaries for our virtual machines, this new capability opens up possibilities for mobility that simply weren’t possible before.
Any Workload, Any infrastructure, Any Time
Our vision at VMTurbo is controlling any workload on any infrastructure at any time.
With the latest release of VMTurbo’s Operations Manager, our customers now have the ability to utilize cross vCenter Migrations to merge clusters in different datacentres. By taking advantage of this brand new feature, we can open up your datacentre and free your workloads in ways that simply were not possible a few months ago.
Our Economic Scheduling Engine is already keeping your environment healthy, but now, you can migrate workloads across clusters on separate vCenter instances, the actions VMTurbo executes are now based on a less constrained set of buyers and sellers. This increases the efficiency gains as our software matches supply and demand of resources from much broader resource pools.
By opening up the market, just like in a real market, this leads to an increase in flexibility. It’s like a brand new Free Trade Zone, your virtual machines are all of a sudden free to live across even more infrastructure, guaranteeing performance and driving efficiency across even more of the datacentre.
The great thing about cross vCenter vMotions is that they can be executed whether you have shared storage or not, so called “Shared Nothing Migrations”. Where shared storage exists, VMTurbo gives you visibility as to which Virtual Machines are running on specific datastores, regardless of which vCenter datacentre they reside on, something you cannot get from the current version of vSphere client.
So when might you want to use this feature?
- Datacentre consolidation: Looking to consolidate 2 or more datacentres? An example would be when a company is acquired (or is merging with another organisation), and the IT infrastructure is to be combined. Using VMTurbo and this new feature, you can migrate the entire datacentre across or use workload placement policies to move the workloads to where you want them to go. VMTurbo will handle maintaining performance and achieve higher Virtual Machine to host densities at the same time.
- Disaster Recovery Planning/Datacentre Load balancing: A lot of organisations now have separate Datacentres for Disaster recovery. Using VMTurbo you can treat both Datacentres as a single entity for resources. The plan view will allow you to test your disaster recovery plans. Or why not load balance across both sites? VMTurbo tears down the boundaries and achieves accurate workload placement and efficiency.
- Migration: When migrating or extending infrastructure, you might need to do a mass migration of workloads for a number of reasons. For example, you might be building a new network fabric (maybe implementing a Software Defined Network), and during the project, you can move your virtual machines around, guarantee performance, while keeping all performance and event history.
The possibilities are endless, and we’re looking forward to being blown away by the creativity of our customers in the near future, as they look to utilize this new capability.