I have been running the Condusiv Technologies V-locity 5 software on a couple of test machines in my home lab over the last couple of days, alongside those VMs I also ran a couple of VM’s with the older V-locity VM product to see how performance differed.
Testing on both products still takes a few days but there are a couple of differences in how the performance data is presented at the end of the test.
In the previous version you ran a Benefit Analysis Report to show you before and after information which generated a nice PDF that you could save, this PDF gave you a page full of information that made for very easy reading.
As you can see above there is some interesting information provided by the V-locity VM Benefit Analysis Report.
With V-locity 5 Condusiv have taken away the Benefit Analysis Report and given you the Benchmark Data Report, this is provided via the VMC console and looks like this.
You can break this down on a per VM basis and then send that off to Condusiv for further analysis and they will return an Excel spread sheet with all of the performance data on it that you can use to extract the data that’s important to you.
Now let’s talk about the results from the testing I carried out.
First of all both the V-locity VM and V-locity 5 products greatly improve your Workload, throughput and IOps to your storage, and also greatly reduces the I/O’s to the storage device, the performance difference in some cases is amazing.
In the above images you can see two graphs (1 and 2), these relate to the Windows 2012 R2 Server and Windows 7 machines respectively.
It’s really interesting to see the results difference between both products, with V-locity VM on the Windows 7 VM you’re still doing more I/Os to the drive with V-locity VM running than you are with the VM running V-locity 5 (64,330,117 for V-locity VM compared to 44,019,505 for V-locity 5), you’re also doing more I/Os per GB with V-locity VM than you are with V-locity 5. It’s obvious that the work Condusiv has done with fine tuning V-locity 5 has worked wonders for performance improvements.
What I am seeing with using V-locity 5 is that there are even more performance gains to be had by running the newer smaller version of the agent compared to the older V-locity VM product which already offered some great performance benefits anyway.
One thing that’s clear is that if you have a Windows estate that’s heavily constrained by poor performing storage you’re going to see an improvement when using Condusiv Technologies V-locity 5.
As a product there is always room for improvement but actually most of those improvements aren’t actually agent related but more VMC \ vSphere Web Client related, it would be really great to see a tie in to vCenter, some better reporting tools available to the end user rather than relying on exporting the results and sending them off to Condusiv (don’t get me wrong they turn around the results quickly enough but sometimes you just want that data to hand). One final area that I would like to see Condusiv working towards is agents for other platforms, I spent many years working in the Windows world and appreciate the Windows agent side of things but Linux is huge in the industry, in fact in my company I would say we are 90% linux 10% Windows and that’s a massive amount of the estate that are missing out on the potential performance improvements on offer here.
I like the fact that the agent is able to be deployed via the VMC console as well as third party Systems Management tools and I love the fact that the whole package is just so small.
I can tell you for sure that I will be running V-locity 5 in my homelab moving forwards, if it saves IO’s going to my SSD’s therefore extending the life of those drives then that’s a good thing in my books.