Virtualization Management Blog

What’s New in the August Release of VMTurbo

I wish you could all see what happens day to day here at VMTurbo! Continuing the automotive metaphor borne by our name, it really is pedal-to-the-metal. Product definition is constantly working with customers to learn exactly what our users need and Development takes on those tasks without fear of scale or complexity - after all, we are the company that developed the Economic Scheduling Engine, the powerhouse behind our ability to optimize and capacity plan in a way that cannot be matched by any other technology out there!

This month's release is true to form and embodies a considerable expansion to our capabilities both in terms of breadth of functionality as well as depth of understanding of what is in essence an incredibly difficult problem to correctly and efficiently solve.

No single capability has generated more interest this month than the release of our Hyper-V compatibility. More than just swapping one hypervisor for another, this represents a significant step forward in that we now support not only multiple instances of ONE brand of hypervisor, but that now this is extended to include heterogeneous environments. Now from a single pane of glass, you have visibility and insight into the diverse infrastructure that most companies are now asking their IT staff to manage. Worried about VMWare pricing structures? Sure you are, but know that regardless of how you end up paying for resources, the actions of VMTurbo (through the internal commodity pricing structures) can optimize for whatever scenario delivers the best application performance at the lowest total cost of ownership.

VMWorld is only a couple of weeks away and we look forward to seeing as many of you there as possible - please stop by the booth and have a drink with us on 'beer' night. Of nearly equal importance will be some significant new capability announcements and we look forward to showing you those. That's all a short time in the future, for now let's have a closer look at a few of the highlights from the August appliance.

1. Microsoft Hyper-V support. Connect to multiple instances of multiple hypervisors.

   See, manage and optimize for the entire IT stack in a single pane of glass.

   

2. The Policy Builder tab has been fleshed out to collect together those functions that logically make sense to offer as an 'active directory' style granular table. We now have together:
   
   a. Quality of Service (QoS) system parameters
   b. Remedial system 'Actions'
   c. VM priority assignment
   d. Email notifications for system and infrastructure events
   
   
3. Support for automated VM configuration; vMem, vCPU, capacity, reservations, limits etc in response to VMTurbo performance recommendations.
   
   
4. While we are looking at this page, lets also look at the email notification policy builder. Here we can tailor our notifications to the type of event and its severity.
   
   
5. Just to mix it up - let's throw in a user interface enhancement to make it easier to work with any aspect of the appliance. How about a common color scheme so that CPU or IOPS in one place looks the same as in another! :-)
   
   
6. We can now tell you that VMTurbo now takes into account the overhead of 'swapping'. Think about it, a host with higher degrees of ballooning or swapping is less attractive to a VM due to higher 'memory prices' (remember, everything in VMTurbo's Economic model is a commodity that must be "bought" by the VM - it's not real money and you don't need to know how much it is, that's just how it works!).
   
   
7. Ok, back to the front-line stuff. How about automated Storage vMotion? Check!
   
   
8. Ready for more Geek stuff?
   
   

   VMTurbo now gathers IOPS and Latency performance metrics on a per VM basis for each Datastore AND the performance metrics between each host/Datastore pairing ( again IOPS & Latency). Why?
   
   Because operations such as cloning, Snapshots etc generate traffic on LUNs and the metrics are not associated with any specific VM. Ergo, there are potentially congested LUNs whose root cause are not immediately obvious because their source cannot be traced back to a VM. We now aggregate these metrics at a Datastore level so we can understand where specific LUNs are congested and make recommendations to Storage vMotion to remove bottlenecks. We do this for all types of storage (Block & File).
   
   Cool!
   
   

9. With all these cool new automation capabilities, you will be pleased to know that you can now see the progress of each resolution as its applied
   
   
   
   Cool!
   
   
10. There are way more than 10 new capabilities in this month's appliance, but for now, lets end this list with 2 cool new 'template' capabilities.

Let's say you were looking to evaluate the price/performance benefits of two different vendor-specific hardware solutions. All you need are the performance metrics of their solution right? How about we give you a template library to get you going? We will continue to update and expand this capability, but these will certainly get you started (and you can always add your own!).

And what about adding addition VM loads? Sure we can clone existing loads, but what if you wanted to simulate the loading (and hence the performance charactaristics) of a VM loaded with a combination of applications that didn't currently exist in your inventory? Build one? That could take a while. How about we just give you a library of standard or common configurations? That would really help (and again, you can always build your own).

That should keep you going for a month - and remember, please come and say hello to us in the booth at VMWorld - we are giving away iPad2's and free beer!