Hi Feng,

> My program will grab each url one by one and run my scraper method with 
> the url and it will output an html file.

Isn't that scRUBYt! by any chance? :-) Well, it doesn't matter whether 
it's scRUBYt!, a Hpricot script or whatever similar, your process will 
be kicked to death really fast - or at least that was the situation when 
I last tried. Maybe they improved their policies somewhat - however I 
doubt that dream  would be usable for these kind of things even after that.

Your best shot is slicehost - it's cheap, fast, easy to setup, great 
support and no nagging. Dreamhost is great for blogs, gallery, wikis and 
that sort of stuff (apart from the ocassional screwups and slowdowns). 
However, hands off if you are trying to run a Ruby script there (unless 
it's Rails or something they accept).

Just my 2c... let me know if you have a different experience.

Cheers,
Peter
___
http://www.rubyrailways.com
http://scrubyt.org