What could your IT department do with 174 days?

Topics in this article

The time and risks associated with deploying virtual infrastructure

This blog is part one in a two-part series on preintegrated systems

When new infrastructure is needed for virtualization and cloud projects, often companies take the customary do-it-yourself route of design, procure, implement, configure, pre-release testing, and production and maintenance. To get a better understanding of this process and the time it requires, plus the possible risks, let’s walk through a scenario for our fictitious business, Company X.   

The sales department in Company X comes to the IT department with a request for virtual desktops. Once that request is in, the IT department takes the following steps:  

Design: The first step in the process is figuring out the requirements needed for the infrastructure. How many virtual desktops need to be supported? How much memory will each require? Once IT has gotten all the answers to their questions, they will design and architect the infrastructure and send it out for peer review.  

  • Duration: 45 days, 200 + working hours
  • Risks: Some of the risks that can come up are inadequate capacity, reliability and scalability. And even if they’ve gotten everything right for this request, there lacks flexibility for future needs. Plus, since the IT department doesn’t put up virtual infrastructure every day, they may lack the best practices knowledge and expertise for the project.

Procure: With the architecture agreed on, IT gets the RFIs, RFPs and POCs needed.

  • Duration: 30 days, 50+ working hours
  • Risks: Since IT departments have options on who to get their infrastructure from, they may have to negotiate multi-vendor contracts. And someone will have to manage and coordinate the delivery of all the parts.

Implement: While waiting for the items to be delivered, the team will project plan and space plan for the equipment. And once the equipment has arrived, it’s time to rack and stack.

  • Duration: 30-60 days, 100+ working hours
  • Risks: Of course, sometimes shipping isn’t always reliable, so the equipment may arrive late; or if it’s from different vendors, the delivery schedules won’t be coordinated. And the project itself can be delayed due to other priorities. If everything does stick to the schedule, the IT department may lack documented best practices and consistency if it’s been a while since they’ve been through the process.  

Configure: Once the hardware is ready to go, IT will deploy and tune software, and flash systems to the latest code.

  • Duration: 15-25 days cycle time, 60+ working hours
  • Risks: The hardware and software may have interoperability issues. And since IT is doing this process manually, there may be user errors, which cause time delays.

Pre-release testing: Before the infrastructure is completely ready-to-go, IT will do user acceptance testing and process integration.

  • Duration: 14 days cycle time, 40+ working hours
  • Risks: If issues are uncovered during this time, they can not only delay the process, but stop it entirely. Plus there may be a lack of reliability and SLAs to test against.

Production and maintenance:  Once the infrastructure is up and running, there will be on-going provisioning, future scaling and maintenance and other updates.

  • Duration: Ongoing
  • Risks: Management of the infrastructure may be through multiple tools. For support needs, there may be multi-vendor support. Plus, there will be ongoing patches and upgrades.

Phew, so for these six steps, not including ongoing maintenance, the IT department at Company X took 134-174 days and 450+ working hours to get the infrastructure up for the virtual desktops. These numbers come from conversations we’ve had with customers, and are not the exception. In an IDC Survey (from July 2011), they said around 40% of the companies they surveyed spend between 3 to 6 months to deploy new IT services.     

From the total time and resources the do-it-yourself process takes, you can see why many companies are turning to preintegrated systems for a faster, more efficient and lower risk approach. Next week we'll share more on Dell's offering in this area – vStart – and what differences and experience Dell brings to these pre-engineered systems.

About the Author: Lani Dame

Topics in this article