aboutsummaryrefslogtreecommitdiffhomepage
path: root/ARCHITECTURE.md
diff options
context:
space:
mode:
authorChocobozzz <florian.bigard@gmail.com>2015-06-09 17:41:40 +0200
committerChocobozzz <florian.bigard@gmail.com>2015-10-29 23:14:54 +0100
commit8c308c2bf7f658945d80be9d5880361238635f5b (patch)
tree2130ae60af58e59dab3df07a5d5cdd5174f91ae8 /ARCHITECTURE.md
parent8cb4b4e00ee57eb98dfe1455b6d2de36fc561797 (diff)
downloadPeerTube-8c308c2bf7f658945d80be9d5880361238635f5b.tar.gz
PeerTube-8c308c2bf7f658945d80be9d5880361238635f5b.tar.zst
PeerTube-8c308c2bf7f658945d80be9d5880361238635f5b.zip
Spawn
Diffstat (limited to 'ARCHITECTURE.md')
-rw-r--r--ARCHITECTURE.md76
1 files changed, 76 insertions, 0 deletions
diff --git a/ARCHITECTURE.md b/ARCHITECTURE.md
new file mode 100644
index 000000000..3cc559fe6
--- /dev/null
+++ b/ARCHITECTURE.md
@@ -0,0 +1,76 @@
1# Protocol (WIP, help wanted)
2
3## Vocabulary
4
5 - Network: several servers communicating each others with this software compose a network
6 - Pod: a server of the network (inspired from Diaspora, no really signification)
7 - Friend: a pod that communicates with yours
8 - Origin pod: the pod on which the video was uploaded and which is seeding the video
9 - Make friend: the action of a server which will join a network (and so become friend with all pods that compose this network)
10
11## Base
12
13### The first run: join a network and make friends
14 * It will generate a RSA key
15 * It will join other networks by checking the configuration file
16 * If the config file doesn't specify other pods, the network will be composed by this only pod
17 * If the config file specifies one or many pods, the server will ask them the list of the pods in the network.
18 The server will add in its friends list all pods that are in > 50% of all other pods friends list and the pods that are asked for the list. For example if there are the following pods in a network with their following friends list:
19
20 http://pod1.com
21 - http://pod2.com
22 - http://pod3.com
23 - http://pod4.com
24 - http://pod5.com
25
26 http://pod2.com
27 - http://pod3.com
28 - http://pod5.com
29
30 http://pod3.com
31 - http://pod5.com
32
33 It will add: `http://pod1.com`, `http://pod2.com` and `http://pod3.com` because it asks to them the list of their friends. Then, it will add all pods that are in >= 50% of pods friends list so: `http://pod5.com`.
34 * The friend-making operation is mandatory and irreversible (we can't change the network for example, even if you are the only one pod in it)
35 * When the friends list is added, the server will present itself to all these friends with the following informations: its **public RSA key** and its **URL**
36 * In returns, it will receive the list of the videos owned by the pod
37 * All the friends have the initial score of 100 points (can be increased to 1000) which represents the reliability of this friend
38 * If the score reaches 0 the friend is revoked (and blacklisted for the future ?) and the video deleted
39
40### Communications
41 * All the communications between pods are signed, encrypted with a RSA key and a symetric encryption
42 * All the requests are retried if they failed 10 times with a factor of 3 (so it will finally fail in ~ 16h)
43 * If a request fails the score is decreased by 10 points
44 * If a request is a success the score is increased by 10 points
45 * The maximum of points would be 1000 (maybe more or less, depending of the server activity)
46 * A pod which receives a request checks if the signature corresponds to the pod it has in its database. Then, it decrypts the body (or ignores it if the signature is not valid)
47
48### Actions on a pod
49 * A pod is a tracker (websocket) which is responsible for all the video uploaded in it
50 * A pod has different user accounts that can upload videos
51 * All pods have an index of all videos of the network (name, origin pod url, small description, uploader username, magnetUri). For example, a test with 1000000 of videos with only alphanum characters and the following lengths: name = 50, username = 50, url = 25, description = 250 has a mongodb size of ~ 1GB.
52 * After having uploaded a video, the server seeds it, adds the meta data in its database and makes a secure request to all of its friends
53 * If an user wants to watch a video, he asks its pod the magnetUri and the frontend adds the torrent (with webtorrent), creates the video tag and streams the file into it
54 * An user watching a video seeds it too (bittorent) so another user who is watching the same video can get the data from the origin server and the user 1 (etc)
55
56## Ideas
57
58 * A video could have more information (detailed description etc) that are not sent on other pods. The user who wants to see these informations has to ask its pod:
59 user asks its pod -> user pod asks origin video pod -> origin video pod responds with the informations -> user pod responds to the user (and puts in cache the informations ?). We could extend this scheme with other informations (user profile etc)
60 * Redondance: if the origin pod is down, the video is not accessible anymore. We could imagine a redondance between pods that keep seeding the video
61 * Server could transcode the video to lower qualities (cost in CPU and disk space)
62 * Server could seed at the demand: for now the server seeds all the videos but it has two drawbacks:
63 - Seeding has a cost and is a long process
64 - After a restart the server has to reseed all the videos (if it has 100 videos it could be very long!)
65 If this solution is choosen, the frontend has to notify the origin pod that it has to seed the video it has not been seeded already
66 * Avoid friend-making being an irreversible operation and a mandatory action at the first startup
67
68## Wanted but no idea how to implement
69
70 * Avoid URL scheme (http or https): if a server wants to implement TLS (and force it), it will break all the communication links
71
72## Debate
73
74 * Any user can view a video of a pod or he should have an account?
75 * Is an ex-friend should be blacklisted for the future?
76 * Handle API breaks in a network. Major update always breaks the API: we need the entire network update to the same major version. Specify a limit date (2, 3 weeks after the release ?) after which all the updated pod switch to the new version of the API? The non updated pod will be ejected of the network because of they would implement the new API