I understand the basic concept that torrenting means transferring files to and from host computers (seeds) and recipient computers (peers) in small chunks. What I don’t understand is what would make this process faster between many seed computers as opposed to just one computer transferring it’s complete file to another computer.
In: 0
Well, one computer transferring a complete file to another computer is just the way the web and file downloads USED to work.
Problem is:
– the one seed computer (or server back in the day) has to pay the bandwidth costs to upload the whole file – multiplied by however many people download it. Bandwidth used to be a lot more expensive.
– if just one seed has the file, then everyone has to download from that one computer. Depending on upload bandwidth and # of simultaneous upload slots that seed has, maybe only a few downloaders can get the file at the same time. And if one of the downloaders has a really crappy internet connection, everyone else will have to wait a loooong time before their turn to download.
Contrast that with:
– once the 1st downloader has a chunk of a file, now there are TWO seeds with that chunk. So subsequent leechers can get it from two places not just one. And with every new leecher become a seed, with copies of the various chunks, subsequent download of those chunks becomes much faster for new leeches.**
– if I am not the only seeder, someone’s download of the file isn’t going to cost my bandwidth the full hit for the file – the cost of uploading is spread out amongst all the seeds.
** this is what the tracker does – when your torrent client has pieces of the file, it tells the tracker “hey. I’ve got part of this file.”. Then, when another client asks the tracker “who has pieces of this file?”, the tracker points that client to you; that client then asks yours what pieces you have and can it have some?
Latest Answers