77d6e9e5cb
Use new missinggo iterator style, and speed up torrent.connHasWantedPieces()
2016-02-07 01:22:31 +11:00
19a304b8ea
connection.Close must be called under the Client lock
2016-02-02 00:45:26 +11:00
afa264e6c5
Reintroduce connection piece inclinations, and begin caching piece priorities
2016-02-01 21:11:41 +11:00
3ed628356b
Use missinggo.Event for connection closing event
2016-02-01 21:08:52 +11:00
b90dacd324
Abstract the pendingPieces as a bitmap
2016-02-01 01:46:28 +11:00
ae9bba2611
Fix downloading of unwanted chunks, and write out downloaded chunks using the connection loop goroutine
...
Doing the writes with the connection loop goroutine prevents memory use blowing out when chunks aren't written out quickly enough.
2016-01-28 05:54:48 +11:00
0ad4dda9fc
Move fillRequests and replenishConnRequests into connection.go
2016-01-24 15:21:17 +11:00
06445f2a1e
It's working and the tests are usually passing
...
I still need to handle "prefetch"-style downloading, and some functions haven't been committed to force this issue.
2016-01-18 18:35:14 +11:00
09050ff2c5
Some missing hunks that fix build
2016-01-04 22:56:36 +11:00
3c446086af
Add Have and Bitfield connection methods
2016-01-04 22:37:49 +11:00
58c997210e
Track peer sources with a dict, and don't exceed high water mark
2015-09-28 15:30:13 +10:00
53259397fc
Initialize connection piece priorities lazily, and using a pool
2015-09-26 17:27:35 +10:00
54339a30e3
Drop connections that contributed to pieces that failed to hash
...
There are peers out there that seem to intentionally spam bad data, and the same pieces will often fail continuously. This really puts a bee in their bonnet.
2015-08-04 01:12:09 +10:00
1cc50f85a6
connection.writer wasn't working efficiently
...
Because of the pseudo-random selection of a communication in a select statement, flushing was occuring prematurely. Also the buffer was needlessly large. For large messages, like outgoing pieces, it's probably better they just go straight to the wire anyway.
This change will make it easier to implement sending of pieces that doesn't block control messages in the future. Any time that the buffer is empty, we'll be able to signal that it's an optimal time to send a piece.
2015-08-02 20:39:31 +10:00
7ecc8108bf
Track connection writes and flushes
...
Helps determine efficiency of the routine.
2015-08-02 14:30:33 +10:00
06dede5690
Track the connection flags for completed handshakes
2015-08-02 04:06:22 +10:00
58dd6ddf93
Implement Stringer on connection
2015-06-29 19:37:52 +10:00
859129fb28
Print peer IDs as escaped ascii
2015-06-22 19:46:26 +10:00
ced5733c88
Improve uploading/seeding
2015-06-16 16:57:47 +10:00
a2df9f8b01
Miscellaneous changes
2015-06-03 00:03:43 +10:00
21fb4933fc
Change the way piece state is exposed to give more detail
2015-06-01 18:22:12 +10:00
bc4aa06c91
Comments
2015-06-01 18:17:14 +10:00
0d13293d71
Support seeding
2015-05-15 08:39:53 +10:00
c5208569ca
Apply sortimports
2015-04-30 00:31:34 +10:00
526d9d738e
Rewrite imports to local bencode and metainfo
2015-04-27 14:55:01 +10:00
28b299e7c8
Rework the Torrent Reader interface, to allow reader options, and add "responsive" as one such option
...
Had several weeks of testing. Removes a lot of the "helper" reading methods, but this was necessary to allow per-Torrent reading options.
2015-04-14 23:59:41 +10:00
60d8ea75a2
Prevent bad metadata_size in extended handshakes from stalling completion of metadata
...
Occasionally bad peers send ridiculous or incorrect metadata_size in their handshakes. If the first acceptable size is wrong, and too small, we'll keep failing metadata completion. If it's too large, honest peers will never send us the pieces we're asking for and we'll never complete the metadata. Now we just adjust the expected metadata size, and keep retrying until we finish. Additionally, we can now request metadata after sending initial messages.
2015-03-27 15:36:59 +11:00
3aa78763aa
Use rogpeppe's sortimports to fix this goimports ordering madness
2015-03-26 17:18:08 +11:00
842a32ae0c
Rewrite import paths for migration from Bitbucket
2015-03-20 16:37:44 +11:00
30c67caced
Allow some overlap between piece prioritization classes
...
This probabilistically improves download speeds when doing readaheads
2015-03-20 10:52:55 +11:00
466b5f0623
Improve connection status info, especially flags
2015-03-18 18:38:29 +11:00
a7dddd9be6
Rewrite handshaking and connection management
2015-03-18 18:28:13 +11:00
203da0aab0
Tidy up extension bytes handling; don't close conn from handshake writer; force protocol encryption for now
2015-03-13 06:21:13 +11:00
e85b7e228b
Add support for Fast Extension
2015-03-12 20:06:23 +11:00
92e0679b04
Change piece prioritization
2015-02-21 15:02:31 +11:00
ae7261c392
Break active peer status line up
2015-02-10 00:19:56 +11:00
2c48152b10
Code comments
2015-01-22 00:42:03 +11:00
8e7e551d76
Fix comment
2015-01-11 00:17:25 +11:00
bcebf569b8
Favour earlier pieces
2015-01-11 00:17:18 +11:00
d54c5ddf24
Remove a bunch of dead code
2014-12-28 12:51:09 +11:00
8d87bb83a6
Refine connection status flags
2014-12-09 00:26:06 -06:00
a4dc96bb12
Make some overlap of piece priorities
...
The goal is to reduce duplicating requests on higher priority pieces.
2014-12-06 21:18:17 -06:00
3c41002317
Set a low water mark for replenishing requests
2014-12-05 00:58:43 -06:00
d58ca65a9c
pieceordering: Rename RemovePiece->DeletePiece and complete tests
...
Note that tests currently fail due to bug in skiplist?
2014-12-05 00:58:04 -06:00
b6ae4caa6a
Set priorities ahead of the current read position
2014-12-05 00:56:28 -06:00
91c2c1f5c7
Piece priorities, torrent read interface and many fixes
2014-12-03 01:07:50 -06:00
7228e40c95
Add piece ordering per-connection piece request priority
2014-12-02 18:22:38 -06:00
78ef36066d
Give each connection its own piece priority order
2014-12-01 03:36:25 -06:00
0f54c2f79e
Use a buffer to reduce syscall.Write calls
2014-11-21 00:04:07 -06:00
e1f720c770
Triage a possible bug blocking while closing UTPConns
2014-11-18 21:53:57 -06:00
Matt Joiner