what am i doing wrong here ??

[dr jones]

iam currently seeding from 2 different trackers

when i seed from tracker a

it connects to 20-40 peers

my upload speed is 0 or very slow

when i seed from tracker b

it connects to 10-30 peers

my upload speed is super fast

tracker b connect to less peers then tracker a does

why is tracker b upload speed is faster then tracker a

tracker a is connected to more peers then tracker b

why is tracker a upload speed slower then tracker a from very slow to 0

those number are just examples

what am i doing wrong here ?

[apextwin146]

tracker a might have partial seeders

[users]

is this really a question??.do you even know how private trackers work?.

[apextwin146]

is this really a question??.do you even know how private trackers work?.

do you really know how to respond to queries?
if you think he dsnt knw then educate him ..

[rvt]

How many other seeds are there on tracker a compared to tracker b?
What's the average completion rate on tracker a vs b?
Is the file a pack, or likely to have people only downloading certain files from the set?
What's the average download speed of users from each tracker?

Any one of those will affect your upload speeds.
If there's hundreds of other seeds, peers may not need your bandwidth.
If they are sharing effectively between the peers (higher avg completion) they wont need seeds so much.
If most of the peers are only downloading a few files from a pack, they will reach a set percentage completed then stop downloading. They will show as peers rather than seeds.
If the trackers have users in different geographical locations, the average download speed could be a lot slower for a tracker where the majority of users are on dial-up compared to one where the majority are on broadband.

[apextwin146]

How many other seeds are there on tracker a compared to tracker b?
What's the average completion rate on tracker a vs b?
Is the file a pack, or likely to have people only downloading certain files from the set?
What's the average download speed of users from each tracker?

Any one of those will affect your upload speeds.
If there's hundreds of other seeds, peers may not need your bandwidth.
If they are sharing effectively between the peers (higher avg completion) they wont need seeds so much.
If most of the peers are only downloading a few files from a pack, they will reach a set percentage completed then stop downloading. They will show as peers rather than seeds.
If the trackers have users in different geographical locations, the average download speed could be a lot slower for a tracker where the majority of users are on dial-up compared to one where the majority are on broadband.

can you elaborate on that ? Lets suppose that there are in total 13 entities 3 seeders and 10 leechers(none partial ) and all on broadbrand ..
If the Avg completion on each of the leecher si ~70% then they would be sharing more amongst themselves?(keeping other factors like geo. etc aside)

[rvt]

They would, but they would also be pulling it from everyone a lot faster.
Say their broadband is 10mb/s download and they are connected to 10 peers/seeds. They can pull 1mb/s off each person.

The same with a 56kb/s dialup would be able to pull only 5.6kb/s from each.

[apextwin146]

But wouldnt it depend on the pieces of the file each of those 10 leechers have?

[dunson]

Are the peers in the faster tracker closer geographically? I find that my slow seeding is usually intercontinental, but my connection gets used well when it is someone within a few hundred miles. Check the IPs on the peer list and see what's up.

[rvt]

Sorry, I've just realised you bolded a part of the post. Thought you were talking about broadband vs dial up (it's been a long day).

Yes, it does depend on the pieces that each peer has for things to be flying along without seeds.
In most clients, they will have a number somewhere, possibly near the number of seeds you are connected to. It'll usually be a float (decimal) number, something like +7.45
This is the number of complete copies that can be made if all the seeds went away.

This is not purely average completion, but it's related.

A number of 10.0 would mean that you can build 10 copies from all the pieces. There would be enough peers with enough of the pieces for your client to work effectively without seeds.
If the number of peers outweighs the number of seeds, and the seeds have full connection queues (they are busy sending to others), your client could still pick up some decent speed using only peers.
A higher average completion in this instance means more copies available from peers.

A number of less than 1 means that you will be connecting to seeds, or, if the queues are full, waiting on other peers to receive and pass on those pieces. You want a higher average completion in this instance as well, as it means you can quickly catch up to the rest of the group from peer connections alone, as they will be sharing very little data between them and so have a lot of open slots.
Once you've caught up, you'll be stuck in the same traffic jam waiting for the few seeds to pass on the pieces. So the higher the average completion, the less pieces you have to wait for once you've caught up.

If you are a seed in the first situation, depending on your upload speed peers may be able to find faster downloads from another peer (or ten).

Edit: if the above doesn't make much sense, I'll take another shot after I get some sleep.