I don't know of any websites, but failing the above advice - conference abstracts are basically the same as journal abstracts - Summarise what you wanted to do, how you did it and the results you got - look at papers published in your field and copy the style.
I can only give my opinion based on a science PhD but what I've found is that initially you'll tend to want to keep the main point as a surprise at the end! - Don't!
State upfront what you're doing, why you're doing it and what you found - same as a normal abstract - it helps people when they are reading it because they already know what to look for (on a poster)
It means on the day someone will look at your abstract and say - thats the area I'm interested in, they're using methods i'm interested in and the results are interesting - lets go see how it looks
hope this helps
S