This post is more of a courtesy than a question... I have an SCCM2007R2 installation running on a Windows 2008 Server. All SCCM roles are on this box along with WDS (NOT CONFIGURED!!!) The issue i experienced was painfully slow multicast OSD deployments via SCCM. To make a long story short, this was a network problem and NOT SCCM. All posts regarding slow multicast point to adjusting registry settings related to WDS...This is still a valid step in resolving slowness.. Make sure you set the registry settings for WDS to default to 1Gbps profile and set a multicast range which doesnt overlap standard ones..( i used 239.0.17.100 - 239.0.17.110) but as long as its not widely used, any MC range will work. That being said, I had my WDS, PXE and DP on a single box. I left WDS UNCONIFGURED and allowed SCCM to do its job. I then updated the regitry settings to default to the 1Gbps profile and set my MC range as stated above. Pxe booted the machine and got to the OSD multicast stage...sooooo slow....soooo slow.... I was confident my network was correctly configured for Multicasting so i put all my attention to SCCM....all SCCM logs (SCCM\logs &SMS_CCM\logs) were error free...wonderfull...now what....i tried rebooting, reinstalling all relevant roles, nothing changed..bang bang bang against the wall..i then loaded up the trusty perfmon and loaded the WDS Multicast Server with the folliowing perf counters (Total Data, Total NACK, total Repair). From seartching multiple posts, the ratio to total data and total NACK should be about 100:1....mine was 25:1....somthing was going on that was dropping all of these packets along the route to the receiver...I then put my attetion strictly on the network side I have a Brocade MCT Core cluster and have Brocade distribution layer switches. My config for the cores are as follows [global] ip multicast-routing ip multicast passive router pim ip igmp version2 [interface] ip pim-sparse ip igmp version 2 My config for the distribution layer is as follows: [global] ip multicast passive The network config is as basic as it gets...it turns on PIM at the core and enabled IGMP throught the core and DL switches. I can see mcache data showing the correct MC address SCCM was using...still no dice...perfmon counters were shoing about 25% packet loss. Here's where i think i got lucky....i issued a show tech on the core to get all technical info. During the export, my line cards reset themselves (no good) and the core failed over to standby while it reset itself. Upon coming back to life, i saw 0 NACK from SCCM and multicast OSD finsihed like a cheetah.... IF ALL ELSE FAILS WITH SLOW MULTICAST DEPLOYMENTS, TRY RESETTING YOUR CORE AFTER YOU MAKE YOUR CONFIG CHANGE To recap, these are the settings that worked for me in an multiple vlan environemtn with multicast routing from sender to receiver SCCM 2007 R2 on Windows 2008 Server All SCCM roles on single box WDS installed and NOT CONFIGURED Registry keys updated for WDS (Default Profile and Multicast Range) PIM Sparse Mode enabled on Core Switches IGMP Snooping enabled on all DL switches Permon counters on SCCM WDS to verify 0 NACK...if you get more than a 100:1 ratio on Total Data / NACK then look at your network! If you are confident everything is setup correctly, schedule a reboot for your core network equipment after u make your MC configs.... I Hope this helps somebody as the primary items here that should be of GREAT help is the perfmon counters for WDS Multicast Server (Total Data, NACK, Repair) Good Luck. Ping me if you need help