Thanks for the feedback. As I had posted in a separate thread a couple of weeks ago, I have had it working against IIS (even pre-2.5 did this). The question was more for how well it was supported / intended to work embedded against IIS given the new threading model?

I have currently implemented this with both IIS 6 and 7 environments. It still works. I am simply looking for more feedback as to how _well_ it works. Perhaps my test scenarios are not robust enough to make this determination before I make the recommendation to move this solution toward production. I am leery of being the only guy out here using it in this manner.

As it is now, I have been able to force application pool recycling during long-running operations and it appears that I can successfully connect on the new worker process. Glimmers of hope, but afraid of standing alone.

Regards,
Frank