Same percentage?

[known]

I was wondering, I am downloading a file and the Peers tab shows 58(684). I know that by increasing the number of Peers my download will be any faster but I was wondering why when I go to the Peers tab and sort by "%", there is only 1 100% person and the rest seems to be 27.3%, the exact same percentage as me. I was wondering why there was only 1 person that had a higher percentage then me, I mean like shouldn't there be atleast 2-3 more people over 30%. Oh and im not downloading from the 100% person.

I'm sorry if this is a stupid question or if its already been asked but I have't seen my question or it's answer anywhere. (Might be because I dont know what to call my question. I found out what the peers brackets mean and that having more peers does nto essentially give faster dl speeds in the FAQ page but I didn't see a question that asked if everyone had the same percentage as you)

Oh and like I dont mean the exact exact percentage like 27.34342323523 or whatever, im still dling from some of the peers at the same percentage (but at 0.5 KB/s)

[DreadWingKnight]

That seed isn't playing nice, and there's nothing you can do.

[known]

Oh, what about the same download thing?

[DreadWingKnight]

That means the seed isn't playing nice with ANYONE.

[known]

Oh that sicko, thanks dread

[Switeck]

That seed probably has upload slots set higher than upload speed, so it is trying to upload slower than 1 KB/sec PER person. With a chunksize of 1 MB or greater, what's happening is the seed is uploading tiny pieces of a chunk (possibly the SAME chunk!) to multiple people...but not till someone gets a COMPLETE chunk will any of the people share that chunk with someone else.

End result, everyone manages to get all the chunks everyone else has and is stuck waiting for the single seed to upload 1 complete chunk to a single person so everyone can leech off them. Naturally, everyone will get each NEW chunk the seed gives out before the seed can complete the next chunk...so everyone ends up waiting on the seed.

Math behind this:

How long does a 1 MB chunk take at only 0.25 KB/sec?

BTW, it wouldn't surprise me if the lone seed is a BitComet client.

This is why if you ever want to seed fast yourself, reduce your upload slots and turn on super-seeding if you're the only seed.

(...but leave the "use more upload slots if <90% upload speed" option ticked in case upload slots are slow.)

[Moody]

This is why if you ever want to seed fast yourself, reduce your upload slots and turn on super-seeding if you're the only seed.

(...but leave the "use more upload slots if <90% upload speed" option ticked in case upload slots are slow.)

That is good advice, curious myself what others seeders find produce fastest results, in my position, with super seeding on, an upload speed of 45kB/s and only one torrent, would you configure say 8 slots so 5kB/s per slot or 4 at 10kB/s with the <90% upload speed" option ticked. One other thing, is there any way of favouring a fast uploader in the swarm?

[splintax]

I'd go with 10KB/s per slot. As for favouring fast uploaders, BitTorrent protocol handles that automatically.

[Switeck]

Assume you're the only seed for a torrent with 1 MB chunksize and all the peers have nothing.

I'll refer to the different cases as 5 KB/sec (and 8 upload slots) and 10 KB/sec (and only 4 upload slots).

At 5 KB/sec, about 200 seconds pass and then 8 peers each have a chunk to share. They start sharing immediately.

At 10 KB/sec, about 100 seconds pass and then 4 peers each have a chunk to share (which assuming you're on super-seed mode means 4 DIFFERENT chunks!) Presumeably, the 4 peers start sharing to others immediately. At 200 seconds, you've sent out 4 more chunks for a total of 8 you've shared...plus the peers in the torrent have had up to 100 seconds to get the 4 chunks from the 4 to first get a chunk.

So although both cases share 8 total chunks to the torrent swarm in 200 seconds...in the 10 KB/sec case the other peers have spent half that time swapping around the 4 already-shared chunks while the 5 KB/sec case only JUST got 8 chunks and hasn't had even a second yet to share to anyone else.

In 300 seconds, the 10 KB/sec case has uploaded 12 total chunks and the torrent swarm has had 100 seconds to swap around 8 chunks. The 5 KB/sec 8-slot case has had 100 seconds to swap around its first 8 chunks.

At 400 seconds, the 10 KB/sec case has uploaded 16 total chunks and the torrent swarm has had 100 additional seconds to swap around 12 chunks. The 5 KB/sec case has uploaded 16 total chunks and the torrent swarm has had 100 additional seconds to swap around 8 chunks.

Summing it all up so far, in both cases after roughly 400 seconds 16 total chunks have been uploaded by you to the torrent swarm. But what is interesting is how much time the torrent swam has had to trade the chunks between themselves. In the 5 KB/sec case, 8 chunk had 200 seconds to be traded. In the 10 KB/sec case, 8 chunks had 200 seconds to be traded, plus 100 seconds trading 4 chunks and another 100 seconds to trade 4 more chunks.

Can you see who's going to win this horserace?

[known]

OMG that guy is using bitcomet... that monster

[Dark Shroud]

I'm having this same problem, I've been downloading from this person for almost a month now. At one point there were 27 peers and only one seed. The seed is using µTorrent, he was at ver 1300 before & now he's up to 1400. He said he was on a slow cable connection when I messeged him through the private tracker. But I also know he's seeding a lot of files. And he claimed he doesn't speak English when I tried to ask him about using super-seed mode. All I know is that I 'm going to be seeding those files for awhile. To help everyone.

EDIT: @ known's post, that figures.

[splintax]

If this is on a private site, I would message the admins and get his upload access removed. :/ Shouldn't be people on there that don't know how to use their client.

[Switeck]

I still disagree -- better a weak seed than none at all. However BitComet's normal settings won't even make it a weak seed if it's on an average or below-average broadband connection.

The WORST part of all this is the blissful ignorance. The lone seeder probably uploads till his upload/download ratio is about 2:1 and leaves when he reaches that. But because he's spread his upload bandwidth so thinly between so many peers, there is quite likely not a single virtual copy between all the peers due to wasted data (uploading the same chunk to different people) and people leaving after their download completes because THEIR share ratio is very high due to the long waits.

[Nefarious]

That seed probably has upload slots set higher than upload speed, so it is trying to upload slower than 1 KB/sec PER person. With a chunksize of 1 MB or greater, what's happening is the seed is uploading tiny pieces of a chunk (possibly the SAME chunk!) to multiple people...but not till someone gets a COMPLETE chunk will any of the people share that chunk with someone else.

End result, everyone manages to get all the chunks everyone else has and is stuck waiting for the single seed to upload 1 complete chunk to a single person so everyone can leech off them. Naturally, everyone will get each NEW chunk the seed gives out before the seed can complete the next chunk...so everyone ends up waiting on the seed.

Math behind this:

How long does a 1 MB chunk take at only 0.25 KB/sec?

BTW, it wouldn't surprise me if the lone seed is a BitComet client.

This is why if you ever want to seed fast yourself, reduce your upload slots and turn on super-seeding if you're the only seed.

(...but leave the "use more upload slots if <90% upload speed" option ticked in case upload slots are slow.)

could also happen when seeding a file that is 4MB pieces (i've seed a few of those when the original seeder didn't come back), and with my 8kB/s for upload it was hell (even uploading to only one peer)