Slow compared download speed

[assassino]

Download this using BitComet, then download using uTorrent

http://kickass.to/su...o-t8359131.html

BitComet: D: 107 kB/s

U: 0 to 90 (I tried putting limits)

uTorrent: D: 1 MB/s

U: 30 kB/s

Why ?

Both have two trackers, DHT and Peer Exchange

One found 16 seeds, and bitcomet found 9.

Bitcomet is green light on port forwarding, so as uTorrent

What am I missing ?

Also, on the peers tab, I can see many downloading at very high speed. Most are uTorrent (1 BitComet user actualy. 1.37DHEv2)

It's probably not a client problem, since I download something at full speed yesterday with BitComet

Just want to know what is going on on this case. Maybe the torrent link above helps.

Edited May 21, 2014 by assassino (see edit history)

[The UnUsual Suspect]

I don't have 43gb available on my drive so cannot test that torrent, but in order for all things to be equal in a comparison you'd need the same group of "unchoked" peers available to both clients. In this case we have a very large torrent with few seeds so it's very likely that one of those seeds was a high speed seedbox and it simply had no connections available when you were testing with bitcomet. There are also some rare cases with users who perceive anyone who uses a certain client to be bad people and not allow connections with them (don't laugh, I've listened to people like this more times than I can count), but most likely it was just the peers that happened to be available at the time.

As for why the seed count was different, it shouldn't be as long as the tests are done at the same time and all the trackers are working but in some rare cases on private torrents that have multiple trackers uTorrent will connect to more than one tracker. Allowing this to happen could get you banned from the private trackers so utorrent users should be sure not to run such a torrent. This problem was reported to them

Here: http://forum.utorrent.com/topic/69855-multiple-private-tracker-question/

In cases like this BitComet will only connect to one tracker at a time and will clear the list of peers if it becomes necessary to switch trackers. This is the only way to avoid breaking the rules governing private trackers.

The only way to do an accurate comparison between clients is to do it in a closed system where you have control of all available peers, otherwise you could get drastically different results even with the same client, but in a strictly bittorrent swarm with all things equal bitcomet and utorrent will perform equally, but bitcomet does have the ability to substantially outperform other clients if there are non-bittorrent sources available.

[assassino]

I think I understood.

Now where is this book of rules saying that no one can put their noses on two trackers at the same time ?

I did not know these laws existed...

So, it means that uTorrent is cheating ?

I heard once that BitComet was bad because no one seeded (yeah, right: now it's the client's fault), and then we were constantly blocked.

Is this real ? Can we on a glimpse tell if that is the case here (being blocked) ? Or uTorrent's code allows the user to invade a private tracker ?

Dont think we need much testing:

Been downloading this torrent a long long time (40 Gigs, always below 200 kB/s), all along wondering why these guys are faster.

It was almost finished, and I decided to test uTorrent. On the first minute this thing went to full speed !

Edited May 21, 2014 by assassino (see edit history)

[The UnUsual Suspect]

Now where is this book of rules saying that no one can put their noses on two trackers at the same time ?

I did not know these laws existed...

Is this a serious question?

So, it means that uTorrent is cheating ?

No, it means that utorrent can report false stats if you use it on a private torrent with two trackers

I heard once that BitComet was bad because no one seeded (yeah, right: now it's the client's fault), and then we were constantly blocked.

Is this real ? Can we on a glimpse tell if that is the case here (being blocked) ? Or uTorrent's code allows the user to invade a private tracker ?

It would be exceedingly difficult to prove, but if you look at your peer IP addresses where the fast data is comming from and try to connect to those peers using a different client and are unsuccessful, that would be compelling evidence.

[assassino]

Is this a serious question?

By your tone the answer seems to be "nowhere: each tracker has it's rules".

It means that utorrent can report false stats if you use it on a private torrent with two trackers

I am loving this "false stat" ( the one that boosted my download speed).

It would be exceedingly difficult to prove, but if you look at your peer IP addresses where the fast data is comming from and try to connect to those peers using a different client and are unsuccessful, that would be compelling evidence.

Good strategy. Will try that.

[The UnUsual Suspect]

I thought your first question was a joke when you said "put your nose on two trackers", but it's a common rule with all private trackers that you get banned if you report upload stats that aren't accurate. If you have two trackers and you upload 1gb to each you're going to report 2gb to each and there is no way for any client to know how much went to peers from each tracker. This is a huge protocol violation. uTorrent seems to have taken the position that this won't happen unless you have an idiot using the client, so they have done nothing to prevent it even after the issue was reported to them. In theory they are probably correct, but bitcomet having been extra careful to be 100% protocol compliant has developed a way to avoid this from happening.

With your specific task it appears there are two likely explanations why the speed difference. One is just luck, when you used utorrent you were lucky to find a fast seeder among a handful of slower ones and this seeder wasn't available when you were using bitcomet. The other is that this fast seeder is blocking connections with bitcomet.

By the way, I think bitcomet does have the ability to edit your useragent data so it would be possible to emulate a different client, however this would result in an instant ban on a private tracker so I don't recommend it. The reason this was added was to allow users in countries that block all connections that have a bittorrent clients user agent. At one point bitcomet switched to using a web browser user agent exclusively to avoid this but some private trackers objected to this so they went back to a bittorrent agent but added the option to let the user edit the string if needed. If you have no problem using torrents then this option should be ignored, but an expert user might be able to use it to confirm if the client was blocked by any specific peer, but even if confirmed I can't see that there would be any value to that info... it's not like you could do anything about it.

[assassino]

Thank you for the information.

Was a valid question, disguised as a joke. Will be more professional next time.

One last question:

I don't understand the term "report" on the phrase "I might be banned if I report an unprecise upload"

I must account for any upload I choose to make ? I'm not being paid to upload...

Went after this user agent switching option to find out that BitComet could easily have been banned because someone disguised as BitComet made something wrong...

I could do the same and "turn" my BitComet into uTorrent.

Would be fair, since I'm apparently paying from someone's mistake. But won't do it. Too much trouble.

[cassie]

I don't understand the term "report" on the phrase "I might be banned if I report an unprecise upload"

I must account for any upload I choose to make ? I'm not being paid to upload..

On practically all private trackers you are required to upload a minimum amount back to the site, in order to maintain your sharing ratio (usually 1:1... or, in other words, to give back to the site the same amount that you have "taken"/downloaded). By reporting an upload amount that is not accurate indicates that you're not willing to share the amount that you originally downloaded (commonly referred to as a "hit-and-run"). The private site will see this as a flagrant violation of their Norms and Rules, and this will promptly get you admonished and/or banned. The sames rules applies to trying to "hide/diguise" the name of your true client (it will seem that you are using their site under false pretenses and not in good faith.

Of course you're not being paid to upload...nobody is...it's just common courtesy, when you belong to a community that is based on SHARING.

Food for thought: Think of how you would feel if everyone else adopted the attitude of "since I'm not being paid, I'll just download (leech) what I need and that'll be it..." Do you think that anyone would be seeding anything for anyone else (for example, you) to download?... Torrents would surely have a very short life-span...

[assassino]

Yes, yes,

I always do upload 2:1 for this reason (except when the torrent really wasn't what I was looking for).

I ask these questions because I never downloaded a private torrent. Never knew there was a way for them to know who I am and how much I am uploading. All we are talking about is new for me.

It's crystal clear now. All these rules seem fair.

Thank you both for the support here.

Until next time.

[The UnUsual Suspect]

Torrents wouldn't exist at all of users don't seed. With popular new torrents the releaser uploads the torrent then stops as soon as other seeds appear and the torrent then lives or dies on it's own merits and the willingness of others to share.

With Private torrents the seeding times and amounts are enforced. If you don't do you're part you will be warned, and eventually banned. The advantage to private torrents is it gets rid of all the greedy people who don't want to share and tends to keep the torrents alive much longer and download faster. You also have a more exclusive pool of peers so you're less likely to be joined by anti p2p groups. Some private trackers are exclusive and you cannot join without an invitation from an existing member in good standing, and others have interviews where you must provide references, but some do allow the general public to join.

[assassino]

I don't know what is going on

Maybe from the fact that I'm not an VIP

But it's happening with many (almost all) torrents

All of them are slower that uTorrent

Bitcomet: 100kB/s

uTorrent: 1MB/s (my bandwidth)

(no matter how long I wait to allow it to "settle")

Also, it takes way longer to find the peers

A single blink on uTorrent

Good bye, BitComet

Was good as it lasted.

Edited September 26, 2014 by assassino (see edit history)