Project

General

Profile

Protocol core-core Version 3 » History » Version 27

Greg Burri, 07/09/2009 01:29 PM

1 1 Greg Burri
h1. Protocols
2
3
This page describes the protocols used by Aybabtu for network communication between cores and between core and GUI.
4
5 7 Greg Burri
h2. Core <-> Core
6 1 Greg Burri
7 21 Greg Burri
h3. Notes pour plus tards
8
9 22 Greg Burri
* Le message _HaveChunks_ doit être multicast et la réponse est un datagramme UDP. Il n'est pas possible de maintenir N-1 connexion TCP en permanence ou N est le nombre de peer. (à mesurer). DU coup ce message peut remplacer le _IMAlive_
10 21 Greg Burri
11
12 23 Greg Burri
See source:protos/network_protocol.proto for message details.
13 1 Greg Burri
14 23 Greg Burri
There is no more than one downloading from B to A. A peer which a download is active is setted as not free.
15
16
There is _NbDownloadThread_ thread T. Each thread will choose a free chunk and download it. Here is the thread loop.
17
<pre>
18 27 Greg Burri
F : the list of all unfinished download
19
C : A list of chunk candidates (empty at start)
20
21
While not end of F
22
   - Take the next file not finished in F -> f.
23
   If the hash chunks of a file are unknown :
24
      - Ask the hashes to the source (see the message _GetHashes _).
25
   For each chunk not downloaded or not being downloaded -> c :
26
      If c has a free peer :
27
         - add c to C
28
   If C isn't empty :
29
      - Take the best chunk in C based on peer.downloadRate. If there is more than one best chunk then choose one randomly.
30
      - Grab the lock for the peer
31
      - Request the download   
32 23 Greg Burri
</pre>
33
34
35
36 8 Greg Burri
h3. Parameters
37
38 10 Greg Burri
Most parameters will be stored as constants in their component.
39
40 9 Greg Burri
|_.Name|_.Value |_.Unit|_.Explanation/Comments|
41 19 Greg Burri
| _IMAliveFrequency_ | 0.1 | [Hz] | The message _IMAlive_ is sent each 10s. |
42 14 Greg Burri
| _ChunkSize_ | 33554432 | [Byte] | (32 MB) |
43
| _LanSpeed_ | 52428800 | [Byte/s] | (50 MB/s) |
44 17 Greg Burri
| _TimeRecheckChunkFactor_ | 4 | - | If the time to download a chunk exceeds _TimeRecheckChunkFactor_ * _ChunkSize_ / _LanSpeed_ the peers are rescanned to choose a better one (better download rate). For example : 4 * 32 / 50 = 2.6 second |
45
| _SwitchToAnotherPeerFactor_ | 1.5 | - | During the recheck of the peers to choose a better one the other download rate must be above the current * 1.5. |
46
| _DownloadRateValidTime_ | 1500 / _LanSpeed_ = 300 s | [s] | (5 min). The download rate of a peer is only valid during this time after its last update. After that its rate is set to the maximum. | 
47 15 Greg Burri
| _TimeBetweenChunksUpdate_ | 500 / _LanSpeed_ = 10 s | [s] | Each 10 s a daemon will scan the _NumberOfChunkToUpdate_ first chunks from the first queued downloads to know which peer has which chunk. |
48 20 Greg Burri
| _NumberOfChunkToUpdate_ | 100 * _LanSpeed_ / _ChunkSize_ = 156 | - | The first 156 hash chunks (3 kB) will by send periodically to each peer. See message _HaveChunks_ in the proto file. |
49 19 Greg Burri
| _FindTimeout_ | 15 | [s] | A search request is abandoned after 15 seconds from the start of the request. See message _Find_. |
50 25 Greg Burri
| _GetHashesTimeout_ | 200 | [s] | Timeout for the message _GetHashes_. |
51 19 Greg Burri
| _GetChunkTimeout_ | 10 | [s] | Timeout for the message _GetChunk_. |
52
| _HaveChunksTimeout_ | _TimeBetweenChunksUpdate_ = 10 | [s] | Timeout for the message _HaveChunks_. |
53 1 Greg Burri
| _GetEntriesTimeout_ | 10 | [s] | Timeout for the message _GetEntries_. |
54 23 Greg Burri
| _NbDownloadThread_ | 3 | - | Number of concurrent downloading thread. |
55 26 Greg Burri
| _BufferSize_ | 1048576 | [Byte] | (1MB). This buffer is user when reading or writing a file. |
56
| _TCPSocketBufferSize_ | 33554432 | [Byte] | (32Mo). Used when data are received. See _QAbstractSocket::setReadBufferSize_. |
57 8 Greg Burri
58 7 Greg Burri
h2. Core <-> GUI