Ask the Experts

ISPs face a fundamental question: should they work to limit P2P traffic

on their networks, or embrace P2P, manage it effectively, and learn to

benefit from it. As with most Internet traffic trends, there is inevitability to P2P traffic.

It is doubtful that ISPs can effectively prevent P2P traffic from infiltrating their networks through protocol blocking, or that they can afford to continually purchase more bandwidth to handle increased traffic. When you add in the benefits from P2P for legal file sharing and content delivery, embracing P2P and accommodating for P2P

network traffic seems to be an ISP’s best solution.

However, today many ISPs continue to fight P2P on their networks, and they face constantly changing technical hurdles, customer dissatisfaction and more importantly irreparable damage to their brand.

Based on our deployment experience we see that the 80/20 rule applies to P2P traffic, where 80% of the users are trying access 20% of the content, which means that we really need to have enough storage to accommodate the 20% of the content. We are able to achieve high byte hit ratios, as high as 80% with 4 Terabytes of data. Even if we add another 4 Terabytes of storage, the byte hit ratio does not increase significantly. Nevertheless, the required storage capacity might increase as higher quality video files like HD are being shared in the future, but nevertheless we don’t foresee the storage requirements going significantly higher. Also a Centralized storage architecture allows all the members of the cluster to share the same files avoiding content duplication, utilizing 100% of the storage capacity and thereby boosting cache hit ratio

Contrary to common belief, the tendency of P2P protocols to adopt encryption is not an innate feature of peer-to-peer sharing but rather a direct result of large-scale deployment of bandwidth shaping by ISPs seeking to curb and control their bandwidth expenses incurred by P2P traffic.

The purpose of the encryption schemes, as used by Bittorrent clients and eMule is traffic obfuscation, rather than protection of users and/or content identity.

Even so, encryption support is disabled by default, so that only users negatively affected by P2P blocking and throttling would use it.

All parties involved, from P2P development community to ISPs to content owners contend that encrypted P2P is a negative development, triggering a war of attrition between DPI vendors and ISPs on one hand and P2P community on the other.

Both Bram Cohen, the inventor of Bittorrent, and eMule team had argued against encryption as a step alienating ISPs, defeating P2P caching and embrace of P2P for legitimate uses.

Indeed, most of major DPI vendors have come out with a support for detection of encrypted P2P within 6 months from the first deployment of P2P encryption-based obfuscation in major Bittorrent clients.

PeerApp is addressing the challenge of P2P encryption in a couple of ways.

One is deployment of combined shaping/caching solution in conjunction with leading DPI vendors.

Such comprehensive one-stop solution for P2P traffic management allows ISP to throttle down the encrypted traffic,

while caching the unencrypted traffic at the same time.

Since the primary reason for P2P users to turn on encryption is to increase performance, such policy shall discourage

encryption usage.

Another direction is capturing the mindshare for P2P caching among ISPs and content owners.

As P2P caching increasing gains traction and ISPs continue to deploy and advertise their P2P caching solutions,

the usefulness of P2P encryption is going to fade away.