Squirrel: Decentralized Blockchain P2P Web Cache Internet Protocol System
Back in 2002, Sitaram Iyer, Peter Druschel and Antony Rowstron presented a paper with a decentralized peer-to-peer web cache alternative.
With this protocol, the cache system that we use today would not be used. Instead, web browsers on desktop machines could be sharing their caches in a much more efficient and scalable web cache.
Furthermore, the team has also evaluated Squirrel’s decentralized web caching algorithms which showed a good performance comparable to a decentralized web cache in terms of ratio, bandwidth usage, and latency. Furthermore, it was scalable, self-organizing, and resilient to node failures.
Web caching is usually deployed on dedicated machines at corporate networks and Internet Service Providers. The traditional approach involved using dedicated hardware that is expensive and has very high administrative costs.
As users grow, hardware upgrades will be necessary as it deals with scalability problems. Dedicated web cache servers represent a single point of failure, if it does not work, then it denies the access to cached web content.
Squirrel works in a different way. It facilitates the sharing of web objects between client nodes. And it works in a decentralized manner when the nodes export their cache to other nodes in the network.
This protocol can pool resources from a wide variety of desktop machines and achieve the performance and functionality of dedicated web cache without having do add more hardware.
If there is an increase in client nodes, then, the number of shared resources increases as well. In this way, Squirrel can scale quickly and automatically. But in order for this to happen, Squirrel needs to use Pastry, a self-organizing , peer-to-peer routing substrate as the object location service. Then, it is responsible for identifying and routing to nodes that cache copies of a requested object.
Squirrel can work ideally with 100 to 100,000 nodes. Moreover, the team believes that Squirrel could be a cheaper, low-management, and fault-resilient solution for web caching in large intranets.