login vsi company logo login vsi company logo 250x40


Login VSI Storage Workload: The ultimate proof of VDI storage performance

The Login VSI Storage Workload is designed to simulate, test and benchmark storage systems with a VDI workload, using a real VDI server infrastructure, real desktop sessions with real applications and data access.

The Login VSI Storage Workload is built on industry-standard workloads and helps storage solution vendors to find the maximum capacity of their offering for virtualized desktop environments. Whether you need to know storage capacity for internal testing, R&D, technical marketing, or a proof of concept to support an enterprise sale, storage vendors run into the usual bottlenecks—CPU and memory before saturating storage. With the Login VSI Storage Workload, storage vendors can finally max out their platform without worrying about other factors and reach an understanding of their true storage performance using industry standard measurements.

The Login VSI Storage Workload offers:

  • Real IOPS, by using realistic data, real applications, real infrastructure and minimal synthetic disk IO load
  • Significantly higher IOPS than the regular Login VSI workloads with a similar CPU and memory profile
  • There is no need for a large launcher infrastructure because the Storage Workload runs in Direct Desktop Mode

90% fewer launchers required

The Login VSI Storage Workload does not require a large launcher infrastructure. This saves you time and expense for running tests. For example, when running a 1,500 user VDI test (+/- 10 server blades), you will need 90% fewer launchers (see table below).

  Login VSI Office Worker Login VSI Storage Workload
Desktop VMs 1500 1500
IOPS per user 6-8 25-30
Launchers (VMs) 40 2

Max out your IOPS

The Login VSI Storage Workload generates at least 200% more disk IOPS in comparison with an Office Worker workload during a Login VSI test (where test users gradually log on), and even 300% more IOPS in a steady state test, but requires the same CPU resources as a normal user. In some cases IOPS go up by 500% depending on your test platform.

IO comparison

The following graph shows IO profile comparison between the Office Worker and the Storage Workload, measured with 150 sessions started in 48 minutes. 150 sessions were started. These tests were performed with all non-critical services disabled. Write-intensive services like indexing or antivirus will significantly impact the IO profile.

Note the dip in CPU Utilization and IOs during the end of the test in the storage workload. This happens because the tested storage solution has reached its limit in this test around 140 users. Because of this, the desktop and applications stop responding. Typically, you want to avoid this scenario in regular validation tests.

storage workload average commands sec

Important Note

Before a large-scale storage test is performed with the Login VSI Storage Workload, first determine the VSImax with a Login VSI Office Worker on a single server. By figuring out what the regular VSImax is first, you can prevent a server overloading on CPU or memory by configuring too many storage users. With the Storage Workload, it is possible to determine the maximum amount of storage sessions that should be started on each server in a large-scale test, without hitting CPU or memory bottlenecks.

Note: The Login VSI Storage Workload is designed to validate VDI storage requirements and is not designed for Microsoft Remote Desktop Services or Citrix XenApp scenarios.


The Login VSI Storage Workload is available as an add-on to Login VSI. Contact us for more information.

What our customers are saying

Jamie Vernon - STARTEK

"Login PI enables us to minimize system downtime and deliver strategic business value like never before."

Jamie Vernon, VP of IT and infrastructure at STARTEK 

Randy Groves - Teradici

"As the developer of the PCoIP protocol, we do a lot of our own internal benchmarking to make sure that we’re continuing to improve our protocol. But we can’t actually publish our internal benchmarks because everybody might think that the tests are biased. So when we want to publish results, we use Login VSI. That way, if a customer wants to repeat our results they can.”

Randy Groves, CTO at Teradici

Johnny Blizzard - Health Insurance Organization

"We now have indisputable empirical proof supporting our claims. With Login VSI, we can run our own applications against various hardware platforms before we move to production! Application owners can see how their code impacts their users on our systems and we can present our findings to management for funding increases. We can also advocate for the users when supporting their claims of poor performance, all because of Login VSI."

Johnny Blizzard, Sr. IT Systems Engineer at a Health Insurance Organization

Login VSI, Inc.

300 Tradecenter

Suite 3460
Woburn, MA 01801

Phone: +1 844 828 3693

Login VSI B.V.

De Entree 85
1101 BH Amsterdam
The Netherlands

Phone: +31 20 705 1200