I’ve been a huge fan of Riverbed‘s Steelhead appliances ever since I put a pair in place at JBI to link our Fort Worth and Dallas offices. The technology they use to shorten data communication across WANs borders on Voodoo in my mind, and the difference in transfer times is stunning the first time you run a controlled test. We’re talking about a change from 10 minutes to 15 seconds on a basic file transfer/open test.

Since I’ve left JBI, I had no way to test Vault and a Riverbed together. Thankfully, the folks at Advanced Solutions did. Read more for my take on their results.

I was on the Civil Community, checking out upcoming events when I jumped to ASI’s site to see about a remote lunch and learn they were hosting. Up on the top right, there was a link that caught my eye, the Autodesk Aficionado. As a taker of Cigar Aficianado, how could I not click it?

It seems ASI has added a new blog, and the first entry was titled Vault WAN Performance. Not much explanation on the entry, but a PDF. Thinking I was about the get the standard Autodesk WAN PDF, I clicked anyway. What a great surprise! Scott Wolslager and Tim Sealschott have done a comparison of Vault performance across a simulated T1 WAN connection with two workstations. The test case was done using Inventor (I think,) and Vault 5.0.

The results are about what I would have expected. Simply put, Steelheads kick ass. I was blown away by the comparisons of LAN vs Wan with and without Riverbed. I still have a few questions for these guys, and will be updating this post when I hear back from them. I’m hoping that I can convince them or Riverbed to run a similar test using C3D data instead of the Inventor files.

Maybe someone from Riverbed can loan me a couple of units to test it here? In the meantime, if you’re going to run a remote office, you should seriously consider checking out Riverbed and this great comparison doc.