Abstract: A data caching and distribution method, performed by a plurality of in a linear communication orbit, includes generating a data request by a first machine to request specific data, and passing the data request along a data request path that tracks the linear communication orbit until the request is received at a second machine, in the linear communication orbit, that returns the specific data in response to the data request. The method includes, at a third machine between the second machine and the first machine in the linear communication orbit, conditionally storing the specific data in a local cache of the third machine according to a data caching method.

Abstract: A method of parallel processing by networked devices, in which distributed data in a network is accessed in parallel to provide results. Commands and queries by a user station define a set of such distributed data. The networked devices access the distributed data in portions, individually process each portion, conduct peer-to-peer communication to aggregate results from processing each portion, and individually determine whether and when to report results. Making effective use of network assets (such as processing and storage at each node), making effective use of concurrency of network assets, limiting bandwidth use between network assets and user stations, and limiting use of assets (such as processing and storage) at user stations.

Abstract: In accordance with some embodiments, a computational machine having one or more processors, a local cache and memory receives from a first machine a data request that is used to request specific data. The computational machine determines whether the computational machine stores the specific data in the local cache. In accordance with a determination that the computational machine does not store the specific data in the local cache, the computational machine responds to the data request by passing the data request to a second machine and returning the specific data to the first machine when the second machine returns the specific data to the computational machine. Additionally, the computational machine determines whether to store the returned specific data in the local cache according to a data caching method.

Abstract: In one aspect, machines in a managed network implements a set of rules that cause individual machines to directly interact with only a small number of machines in the network (i.e., a local neighborhood within the network), while the independent local actions of the individual machines collectively cause the individual machines to be self-organized into one or more communication orbits without any global control or coordination by a server or an administrator. The communication orbits are used for supporting network, security and system management communications in the managed network.

Abstract: In one aspect, machines in a managed network implements a set of rules that cause individual machines to directly interact with only a small number of machines in the network (i.e., a local neighborhood within the network), while the independent local actions of the individual machines collectively cause the individual machines to be self-organized into one or more communication orbits without any global control or coordination by a server or an administrator. The communication orbits are used for supporting network, security and system management communications in the managed network.

Abstract: In a network, a set of machines communicate pairwise, each conditionally adjusting messages in response to its own local state, and each in response to statistical methods conditionally propagating those messages, with the effect that problems with that network, or with a subset of its machines, are reported to a receiver/server. Only a substantially constant number of reports are made to the receiver/server, even when there are a substantial number of such machines able to detect that problem. When a problem is reported, a similar technique causes the machines to collectively evaluate and report suggested causes for that problem. Messages are propagated from each machine to another using locally random global locality. The machines in the network, in response to statistical techniques, organize hierarchically in O(log n) time, where n is the number of machines in the network, substantially without any requirement for nonlocal message exchange.

Abstract: A server receives “registration” messages from nodes in a network, sufficiently infrequently to prevent overburden. Registration messages include information about registering nodes, other nodes, or node relationships. The server sends “instruction” messages to registering nodes, requesting information about the network. Each registering node obtains the requested information, and forwards those messages to other nodes, which act similarly. Nodes receiving instruction messages sometimes send “reporting” messages to the server, providing information it originally requested. The server thus receives detailed information about the network, without communicating with too many nodes, and without aggregating too much of that information. The server can interact with users using only resources growing at O(1) with the size of the network.

Abstract: A method of parallel processing by networked devices, in which distributed data in a network is accessed in parallel to provide results. Commands and queries by a user station define a set of such distributed data. The networked devices access the distributed data in portions, individually process each portion, conduct peer-to-peer communication to aggregate results from processing each portion, and individually determine whether and when to report results. Making effective use of network assets (such as processing and storage at each node), making effective use of concurrency of network assets, limiting bandwidth use between network assets and user stations, and limiting use of assets (such as processing and storage) at user stations.

Abstract: In a network, a set of machines communicate pairwise, each conditionally adjusting messages in response to its own local state, and each in response to statistical methods conditionally propagating those messages, with the effect that problems with that network, or with a subset of its machines, are reported to a receiver/server. Only a substantially constant number of reports are made to the receiver/server, even when there are a substantial number of such machines able to detect that problem. When a problem is reported, a similar technique causes the machines to collectively evaluate and report suggested causes for that problem. Messages are propagated from each machine to another using locally random global locality. The machines in the network, in response to statistical techniques, organize hierarchically in O(log n) time, where n is the number of machines in the network, substantially without any requirement for nonlocal message exchange.