Contrast Between The Bittorrent And Gnutella

Published Date: 02 Nov 2017

Bittorrent: Bittorrent is a protocol that facilitates the mechanism of peer to peer file sharing and helps in distributing large quantity of information(data) over the internet. Bit torrent is very popular for transferring huge amount of data. Bram Cohen is the programmer who designed this protocol in april 2001 and the july 2, 2001 is the release date of Bittorrent. Bittorrent uses the resources of peer nodes to effectively distribute large quantity of data. Here the resources of peer nodes means the access bandwidth and disk storage. A torrent include a central constituent known as tracker and all currently active peers. Bittorrent has two types of peers known as seed and leacher. Bittorrent make difference between these different peers depending upon the downloading status; the client that have complete copy of file and continue to serve other peers called seeds and the client that are still downloading the file are called the leachers. The tracker just acts as central constituent of the system and doesnot involve in the actual distribution of the file. However the tracker collects the meta-information about the peers which are currently active and behave as a meeting junction for all the clients associated with torrent.

A user downloads a torrent file usually from a web server which includes the IP address of the tracker and thus user connects with existing torrent. For initialization of new torrent, first of all user requires a web server that allows to find out the tracker and initial seed with a complete copy of the file. For the update of the tracker’s global view of the system, active clients their state to the tracker periodically every 30 minutes or when joining or leaving the torrent. After joining the torrent ,by the help of the tracker, a new client gets a list of active peers to connect to. Usually, the tracker gives 50 peers chosen at random from active peers but the client wants to maintain connections to 20-40 peers. In case, a client doesnot able to maintain atleast 20 connections, then it again contacts the tracker to obtain extra peers. The collection of peers to which a client is connected is known as its peer set.

Bittorent contains both PC to PC sharing and HTTP hosting. Whenever a torrent is created, and that torrent shouldnot be trackless torrent, then the first seeder is on a dedicated web server and addition seeds can be PCs and even more Web Servers. On contrary Gnutella network doesnot have a dedicated web server for purpose of hosting the file. However Gnutella also support partial file sharing like leecheers in Bittorrent.

Bittorrent is great discovery in the field P2P computing. Bittorrent is not totally P2P content sharing protocol. However bittorrent depends on other global components to help users in the network to locate and work in style of P2P network system. P2P system donot scale well during the process of download of large files. Thus this huge problem of P2P system has adverse effect in the world of network computing. The high download speed and very lower upload speed in a internet connection is another hurdle. So in P2P system , when downloading file the source peer bears the grunt of the download because it must appropriate portions of it limited upload bandwidth among the requestors of the file. In this way, traditional P2P system is not scalable and efficient. Therefore, all users have to face poor download rates. To ensure fairness in P2P, bittorrent breaks down the file into small pieces of same length which are then traded between peers . File pieces are downloaded in nonsystematic way from the source sothat peers can exploit the use of peer connection to other downloaders to get file pieces they donot have. In Bittorrent, file pieces are explained in a small meta-info text file that ends with a .torrent suffix. These file pieces are in encoding format that have messages about the file pieces that makes up the original file. Peers download files from the huge collection of files . The global components of bittorrent P2P network are torrent servers, web search engines and the tracker server. The communication between the peers and global components are possible due to simple HTTP request/response messages. The web servers are torrent servers that stores .torrent files. First user create a .torrent file and upload it to a torrent server and finally other peers can obtain it. In this way user can share a files in bittorrent. The web search engines helps peer to locate the .torrent meta data file on the network. Web search engine acts as global directory of available resource in the Bittorrent P2P network. So, they take different queries of user for a .torrent and give back a resultset page of all possible match.

Security issues in bittorrent

Some very important threats that bittorrent encounters are explained below:

Malicious Payload insertion: It is a security threat found in bittorrent p2p network environment. This threat exist due to the large number of users in the swarm.

Selfish Nodes/Free Riders: In case of media streaming, there should be a continuous and an uninterrupted service of streaming content. Selfish nodes intentionally donot pass the streaming content forward in the swarm.As a result there is slow decay of overall system performance due to presence of many selfish or cheating nodes.

Whitewashers Problem: In some reputed p2p system , node has given responsibility for sharing uplink bandwidth and thus some users tend to change their identities. The users cheat participating in the swarm. It means users download the streaming content without contributing to the swarm. Due to this activities , there is negative impact on their reputation. Finally they ran away from the swarm and come again taking another new identity.

DDos attack: It is another security threat that focus on to overwhelm the network bandwidth and donot allow access to the servers. The current streaming technology hasnot implement large effort to abolish these DDos attacks.

Gnutella : The founder of Gnutella are Justin Frankel and Tom Pepper. Gnutella was first developed in 2000 A.D. and release under GPL license. Gnutella is a protocol that gives simple framework for distributing and sharing files over the internet that is scalable and robust. Gnutella uses the principles of peer-to-peer networking to allow users to share data.

Gnutella protocol is run over TCP/IP a connection-oriented network protocol. A typical session of this protocol includes a client connecting to a server. The client then sends a Gnutella packet advertising its presence. This advertisement is spreaded by the servers via the network by recursively forwarding it to other connected servers. All servers that receive the packet reply with a similar packet about themselves. In the same way, queries are propagated with positive responses being routed back the same path. After the specification of resource and selection for downloading, a direct point to point connection is made between the client and the host of the resource and the file downloaded directly using HTTP. In this situation the server will behave as a web server which has capacity of responding to HTTP requests.

In Gnutella, peers in the network are joined by point to point connections to a set of neighbouring peers to establish an unstructured overlay. The Gnutella protocol has five descriptors. They are Ping, Pong, Query, QueryHit and Push. They are explained below in brief.

Ping: In the traditional TCP/IP, ping means that "Are you alive? Query to another host address. But the gnutella ping is for the announcement that the initiating host has joined the network. The servent response with a Pong whenever the servent gets the Ping and also forward the ping to all other servents to which it is connected.

Pong: Its function is just to response to a Ping request. When servent receives ping from the sender, then the servent also reply with pong which contains information consisting of the IP address and port on which the servent is listening, the number of files it is sharing with the network and the total amount of data contained in those files.

Query: It is used for the search mechanism. The query has two important requests: a minimum connection speed and a search string containing the data that the user is looking for.

QueryHit: A servent will respon to a Query descriptor with a QueryHit descriptor if it matches the search string in the Query with one of the files that it is sharing with the network and if it can also achieve the minimum connection speed specified in the query.

Push: The Push descriptor makes the Gnutella network to continue to operate when a servent that is giving files stays behind a firewall that willnot allow direct connections to be established to the host.

Third P2P network system

The example of the first generation P2P network is Napster and second generation is Gnutella file retrieval. The third generation is a hybrid of both first and second generation system. The third P2P network is based on distributed hash tables which is a technology based on hash tables enabling identification and retrieving, in distributed systems like some P2P networks, of information. The whole table is distributed on the network, each node has a part of it. Chord is a typical third generation P2P system which has no centralized control or hierarchical organization. The chord is useful in locating the node that stored a particular data item in a highly scalable dynamic distributed environment. The chord has good lookup performance, and high failure tolerance.