Adaptiva’s simple one-line announcement of OneSite Peer to Peer PXE triggered a storm last month.

Amidst all the craziness, the announcement sent thousands of hits from around the world to my blog.

It also altered the course of several large SCCM deployments, some as far away as Europe and Australia.

Along with all these has come a deluge of proof-of-concept deployments, feedback, feature requests, and a set of fascinating scenarios that people wish to enable using the revolutionary protocol we have created.

In today’s blog I will describe one overwhelmingly requested use case that has emerged as the holy grail for a lot of companies.

The Holy Grail of OSD Scenarios

This is how one admin described their scenario to us:

1. We have an automated build process using SCCM/MDT task sequences
2. It installs the OS, drivers, applications, and integrates with our backend processes
3. It works great, we put in a bare metal machine, and get a fully built production machine out
4. But it works only in the head office – because it needs to download Gigs of data every time
5. We’d like it to work seamlessly worldwide, even over WAN – without changing our build process

People want their SCCM/MDT build process to work anytime, anywhere, and at the same speed as it works in their headquarters, all without building out the enormous amount of server infrastructure that is normally required to support these scenarios.

Is P2P PXE Enough?

As we explored these scenarios, we recognized not only the coolness, but also the enormity of some of these requirements.

Our conclusion was that P2P PXE provides some vital elements of what was needed, but a lot of additional engineering was required to provide the seamless enablement that people were asking for.

Back To The Whiteboards

Having fully defined the use case and identified its detailed requirements, we went back to the whiteboards and carefully crafted the additional superstructure that was required on top of the P2P PXE protocols.

As we developed the code, we also selected a large retail organization to work with. We had already concluded a successful lab POC with them, and they were eager to prove the P2P PXE stack in production, identify any shortcomings, and help validate the product’s new capabilities to make sure they met these severe requirements.

With the release of Version 1.2.86 of OneSite on Friday, after a month of close work with a few carefully selected customers, we’re finally ready with a release that completely enables this demanding scenario.

Simplifying Windows OSD

The resulting solution is simple and elegant. You’re ready to extend the reach of your OSD build process to all corners of the world.

  1. Enable P2P PXE
  2. Modify the boot image for standalone use
  3. Inject the OneSiteDownloader tool into the boot image
  4. Add an action to run OneSiteDownloader in your Task Sequence
  5. Use Adaptiva OneSite Content Push to pre-stage your images and packages

The Results

After we wrapped up the remaining pieces, I had the joyful experience of idly watching by, as two experienced SCCM admins expertly navigated their build process, and successfully deployed a production Windows image to a laptop across the WAN, in the middle of a busy workday. OneSite successfully downloaded all boot images, OS images, and packages locally, while the WAN continued to hum along.

There are far reaching implications for how SCCM is deployed, and how it gets used for deploying production builds.

I’d like to once again thank all our customers who provided brilliant inputs and allowed us access to their production networks to deploy and fine tune these technologies.