Rate based failure detection
First Claim
1. A method for detecting anomalies in a computer network in which a plurality of publishers publish data and a plurality of subscribers subscribe to published data via network routes in the computer network defined to satisfy corresponding expected quality of service (“
- QoS”
) of the individual subscribers, the method comprising;
under control of one or more computing devices configured with executable instructions,receiving a first notice from a first node within the computer network that data received on an incoming connection is not being received according to an expected quality of service from an upstream node;
determining whether a second notice has been received from a second node within the computer network, the second node being configured to also receive data on an incoming connection from the upstream node;
determining, based at least in part on the first notice and the second notice, that an anomaly exists within the computer network when (i) the first and second notices both indicate that data received on the corresponding incoming connections is not being received according to the corresponding expected qualities of service and (ii) a notice from the upstream node indicating that the data forwarded on an outgoing connection is not being forwarded according to another expected quality of service has not been received;
determining whether one or more other notices have been received from other nodes receiving data from the upstream node; and
in response to determining that one or more other notices have been received from other nodes receiving data from the upstream node, indicating that a cause of the anomaly is at least one of the upstream node or a data path to the upstream node in the computer network.
3 Assignments
0 Petitions
Accused Products
Abstract
This disclosure describes, in part, a system management component and failure detection component for use in a power grid data network to identify anomalies within the network and systematically adjust the quality of service of data published by publishers and subscribed to by subscribers within the network. In one implementation, subscribers may identify a desired data rate, a minimum acceptable data rate, desired latency, minimum acceptable latency and a priority for each subscription. The failure detection component may identify an anomaly within the network and a source of the anomaly. Based on the identified anomaly, data rates and or data paths may be adjusted in real-time to ensure that the power grid data network does not become overloaded and/or fail.
20 Citations
19 Claims
-
1. A method for detecting anomalies in a computer network in which a plurality of publishers publish data and a plurality of subscribers subscribe to published data via network routes in the computer network defined to satisfy corresponding expected quality of service (“
- QoS”
) of the individual subscribers, the method comprising;under control of one or more computing devices configured with executable instructions, receiving a first notice from a first node within the computer network that data received on an incoming connection is not being received according to an expected quality of service from an upstream node; determining whether a second notice has been received from a second node within the computer network, the second node being configured to also receive data on an incoming connection from the upstream node; determining, based at least in part on the first notice and the second notice, that an anomaly exists within the computer network when (i) the first and second notices both indicate that data received on the corresponding incoming connections is not being received according to the corresponding expected qualities of service and (ii) a notice from the upstream node indicating that the data forwarded on an outgoing connection is not being forwarded according to another expected quality of service has not been received; determining whether one or more other notices have been received from other nodes receiving data from the upstream node; and in response to determining that one or more other notices have been received from other nodes receiving data from the upstream node, indicating that a cause of the anomaly is at least one of the upstream node or a data path to the upstream node in the computer network. - View Dependent Claims (2, 3, 4, 5)
- QoS”
-
6. A computing system for detecting anomalies in a computer network in which a plurality of publishers publish data and a plurality of subscribers subscribe to published data via network routes in the computer network defined to satisfy corresponding expected quality of service (“
- QoS”
) of the individual subscribers, comprising;one or more processors; and a memory coupled to the one or more processors and storing program instructions that when executed by the one or more processors cause the one or more processors to; receive a first notice from a first node within the computer network; determine whether the first notice corresponds to that data incoming at the first node is not being received according to an expected quality of service; in response to determining that the first notice corresponding to that data incoming at the first node is not being received according to an expected quality of service, determine whether a second notice is received from a second node within the network indicating that data forwarded on an outgoing connection is not being forwarded according to another expected quality of service, the second node being upstream of the first node; and in response to determining that the second notice has been received, determine based at least in part on the first notice and the second notice that an anomaly exists within the network when the second notice has not been received; determine whether one or more other notices have been received from other nodes receiving data from the second node; and in response to determining that one or more other notices have been received from other nodes receiving data from the second node, indicate that a cause of the anomaly is at least one of the second node or a data path to the second node in the computer network. - View Dependent Claims (7, 8, 9)
- QoS”
-
10. A method for detecting anomalies in a computer network in which a plurality of publishers publish data and a plurality of subscribers subscribe to published data via network routes in the computer network defined to satisfy corresponding expected quality of service (“
- QoS”
) of the individual subscribers, the method comprising;under control of one or more computing devices configured with executable instructions, receiving a first notice from a first node within the computer network; determining whether the first notice corresponds to that data incoming at the first node is not being received according to an expected quality of service; in response to determining that the first notice corresponding to that data incoming at the first node is not being received according to an expected quality of service, determining whether a second notice is received from a second node within the network indicating that data forwarded on an outgoing connection is not being forwarded according to another expected quality of service, the second node being upstream of the first node; and in response to determining that the second notice has been received, determining based at least in part on the first notice and the second notice that an anomaly exists within the network when the second notice has not been received; determining whether one or more other notices have been received from other nodes receiving data from the second node; and in response to determining that one or more other notices have been received from other nodes receiving data from the second node, indicating that a cause of the anomaly is at least one of the second node or a data path to the second node in the computer network. - View Dependent Claims (11, 12, 13, 14, 15, 16, 17, 18, 19)
- QoS”
Specification