Category: Technology

Recovered from the Wayback Machine.

I checked out Circle as well as Chord as P2P networks. These are excellent efforts and should be note to anyone who is interested in P2P systems. As with KaZaA, much of the P2P cloud is transient and located on the peers themselves. The folks at Userland should look at how this can be done with Radio 8.0 if they want a true, distributed backend to the product.

I have a feeling the cloud part isn’t the issue — it will be the Radio backend and this assumption of one controlling application per weblog. At least, that’s what I found when I started peeking around a bit. Perhaps folks more knowledgeable about Radio will have a better idea.

Back to the P2P systems: aside from a key entry point (and all of these systems need this and there’s a reason why) the P2P clouds are without iron. Aside from the key entry point.

Why is the entry point needed? Because each P2P circle is too small (yes it is) to make it efficient to send a bot out into the open Internet, knocking at IPs looking for a specific node of any one of these systems. All P2P systems are too small for this to be effective, Napster, Gnutella, and so on. Think about it — how many nodes are online now in the Internet? I wouldn’t even try and guess the number but I imagine millions and millions. Now you have a P2P network with about 200,000 nodes. Needle in haystack. Right?

Well, not necessarily. Depending upon the dispersion level of the nodes of the P2P network, it might not be that difficult to find an entry node into the network. So with a bot and a handshake protocol implemented at each node you could have a golden gateway — an entry point totally without iron.

However, the problem with this approach is you then have to have a bot for every system you want to join: Groove, Gnutella, Circle, and so on. What a pain.

Wouldn’t it be better to have all these systems provide a common form of identification as well as a common handshake and direction protocol and then have one type of bot that’s smart enough to tap on the door of the nearest P2P system and say “I’m looking for so and so”? And wouldn’t it be better to have each system learn about the others when contacted, such as when a bot returns to a node with a connection into Circle, it also happens to have information about the nearest golden gateway node to Gnutella?   And would it be such a resource burden to have the node check every once in a while to make sure it’s neighboring nodes are still online? So that when our bot of discovery comes calling, it’s given up to date information?

What’s the cost of a ping?

You know, I have so many bots crawling my servers that I’m amazed it’s still standing at times. But none of them work together. If they did, and if they were smarter, and if our sites had something a bit smarter than just open ports, firewalls, or web servers — then maybe we could do without DNS and centralized repositories of information such as UDDI.

Just some more grand ideas to throw out and see if people think I’m full of little green beans again.

Recovered from the Wayback Machine.

In reference to the last posting, Julian mentioned that perhaps Kazaa and it’s supernodes have more of an aluminum core because the cloud that supports the Kazaa P2P network is still mallable — the Supernodes that provide the cloud services are fluid and can change as well as go offline with little or no impact to the system.

I imagine, without going into the architecture of the system, that more than one Supernode is assigned to any particular subnet, others to act as backups, most likely pinging the primary Supernode to see if it’s still in operation. Out of operation, the backup Supernode(s) takes over and a signal is sent to the P2P nodes to get services from this IP address rather than that one. The original Supernode machine may even detect a shutdown and send a signal to the secondaries to take over.

Or perhaps the Supernode IPs are chained and the software on each P2P node checks at this IP first and if no response occurs, automatically goes to the second within the Supernode list and continues on until an active Supernode is found. This would take very little time, and would, for the most part be transparent to the users.

Again without access to any of the code, and even any architecture documentation (which means there’s some guesswork here) the algorithm behind the Supernode selection list looks for nodes that have the bandwidth, persistent connectivity, and CPU to act as Supernodes with little impact to the computer’s original use. The member nodes of each KaZaA sub-net — call it a circle — would perform searches against the circle’s Supernode, which is, in turn, connected to a group of Supernodes from other circles so that if the information sought in the first circle can’t be found, it will most likely be found in the next Supernode and so on. This is highly scalable.

So far so good — little or no iron in the core because no one entity, including KaZaA or the owner’s behind KaZaA can control the existence and termination of the Supernodes. Even though KaZaA is yet another file sharing service rather than a services brokering system, the mechanics would seem to meet our definition of a P2P network. Right?

Wrong.

What happens when a new node wants to enter the KaZaA network? What happens if KaZaA — the corporate body — is forced offline, as it was January 31st because of legal issues? How long will the KaZaA P2P network survive?

In my estimation a P2P network with no entry point will cease to be a viable entity within 1-2 weeks unless the P2P node owners make a determined effort to keep the network running by designating something to be an entry point. Something with a known IP address. Connectivity to the P2P circle is the primary responsibility of a P2P cloud. KaZaA’s connectivity is based on a hard coded IP. However, small it is, this is still a kernel of iron.

We need a way for our machines to find not just one but many P2P circles of interest using approaches that have worked effectively for other software services in the past:

We need a way to have these P2P circles learn about each other whenever they accidentally bump up against each other — just as webloggers find each other when their weblogging circles bump up against each other because a member of two circles points out a weblog of interest from one circle to the other.

We need these circle to perform a indelible handshake and exchange of signatures that becomes part of the makeup of each circle touched so that one entire P2P circle can disappear, but still be recreated because it’s “genectic” makeup is stored in one, two, many other circles. All it would take to restart the original circle is two nodes expressing an interest.

We need a way to propogate the participation information or software or both to support the circles that can persist  regardless of whether the original source of said software or information is still operating, just as software viruses have been propogated for years. Ask yourselves this — has the fact that the originator of a virus gone offline impacted on the spread of said virus? We’ve been harmed by the technology for years, time to use the concepts for good.

We need a way to discover new services using intelligent searches that are communicated to our applications using a standard syntax and meta-language, through the means of a standard communication protocol, collected with intelligent agents, as Google and other search engines have been using for years. What needs to change is to have the agents find the first participating circle within the internet and ask for directions to points of interest from there.

Standard communication protocol, meta-language, syntax. Viral methods of software and information propogation. Circles of interest with their own DNA that can be communicated with other circles when they bump in the night, so to speak. Internet traversing agents that only have to be made slightly smarter — given the ability to ask for directions.

Web of discovery. Doesn’t the thought of all this excite you?

Recovered from the Wayback Machine.

A true P2P cloud does not have a core of iron. By this I mean that there can be no static IP or server providing the gateway or facilitating the communication between nodes within a distributed application.

You can argue this one with me for years and you won’t convince me otherwise. I know that Groove has an iron core cloud. I know that Userland is thinking of an iron core cloud that can move about the nodes. UDDI is based on the premise of a centralized source of information about services that just happens to get striped and mirrorer. Striped — chunked off. Mirrored — distributed to different servers. And don’t focus on the the distributed in the latter, keep your eye on the server.

Server == iron

iron == control

Freenet comes closest to being the truest form of a cloud but there is an assumption that the gateway to the cloud must be known in some way, a pre-known entrance. According to the Ian Clarke’s Freenet: A Distributed Anonymous Information Storage and Retrieval System, “A new node can join the network by discovering the address of one or more existing nodes through out-of-band means, then starting to send messages”.

Can we have P2P clouds without some touch of iron? Can we have transient gateways into P2P networks without relying on some form of pre-knowledge, such as a static IP?

Ask yourselves this — I’m looking for information about C#, specifically about the CLR (Common Language Runtime) and the Common Language Interface (CLI).

Keys are: C# CLR CLI

Go to Google, enter the words, click on I’m Feeling Lucky — and say hi to me in passing.

We don’t need P2P clouds with cores of iron; what we need is new ways of looking at existing technologies.

to be continued…