I read this thread; http://social.technet.microsoft.com/Forums/en-US/configmgradminconsole/thread/265451b3-0cc8-49b3-90b5-95228c546a02 And this post; http://blogs.technet.com/b/configurationmgr/archive/2010/03/22/clarification-on-issues-resulting-from-the-use-of-supernets-in-configmgr-2007.aspx I'm sure that both of the SCCM sites in our forest are using a AD site boundary that is a supernet. There was some overlap that started my research into Boundaries. Our AD structure is politically plagued and is not clean at all. Our subnetting is based on location and has every possible configuration /21 /22 /23 and with public and private ranges in use as well. I hope I can get some clarification on a method to resolve this issue. We do have excellent documentation of the subnetting altho there are up to 100 subnets listed. So is our best plan to remove the AD site Boundaries and add a subnet entry for each subnet in both sites? Or should we use IP ranges?

Hi, I have been using IP ranges in a similar setup. AD sites won't work with Supernets but my IP ranges work like a charm.Kent Agerlund | http://scug.dk/members/Agerlund/default.aspx | The Danish community for System Center products

I concur with Kent, IP Ranges are best for convoluted scenarios like this. Note that subnets will also cause discovery problems in your scenario because AD Discovery directly queries the subnets defined in AD for assignment and because of your subneting, the same problems will happen.Jason | http://myitforum.com/cs2/blogs/jsandys | http://blogs.catapultsystems.com/jsandys/default.aspx | Twitter @JasonSandys

So are you using IP ranges based on subnets to avoid creating supernets with your ranges? I have 55 subnets with unique IP ranges, do I need create a mSSMSRoamingBoundaryRange for each? EDIT: Also, is there a perference that works better with imaging/OS deployment?