Methods and systems to operate an Ethernet ring protection switching (ERPS) network. In one embodiment, a method performed by a network device in the ERPS network device is disclosed. The ERPS network comprises a major ring and a sub-ring, where the network device is a first interconnection network device that is on both major ring and sub-ring. The method comprises determining that a topology change notification is received from the sub-ring, identifying a link that is coupled to the network device and that is towards a second interconnection network device, where the second interconnection network device is also on both major ring and sub-ring. The method further comprises terminating the topology change notification and transmitting a first message using the link towards the second interconnection network device, where the first message is a filtering database (FDB) flush message.

A system and method of identifying faulted devices in an Ethernet network of a building management system (BMS). The system includes a network manager device, a first client device, and a second client device. A first port of the first client device is connected to a first port of the network manager device, a first port of the second client device is connected to a second port of the first client device and a second port of the second client device is connected to a second port of the network manager device. A third port of the first client device is connected to a port of a device external to a ring of the network. The network manager device is configured to transmit a frame having a count field.

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for three-dimensionally stacked neural network accelerators. In one aspect, a method includes obtaining data specifying that a tile from a plurality of tiles in a three-dimensionally stacked neural network accelerator is a faulty tile. The three-dimensionally stacked neural network accelerator includes a plurality of neural network dies, each neural network die including a respective plurality of tiles, each tile has input and output connections. The three-dimensionally stacked neural network accelerator is configured to process inputs by routing the input through each of the plurality of tiles according to a dataflow configuration and modifying the dataflow configuration to route an output of a tile before the faulty tile in the dataflow configuration to an input connection of a tile that is positioned above or below the faulty tile on a different neural network die than the faulty tile.

A bus coupler for a network, in particular for an optical ring network, includes: a bus participant interface for data connection to at least one bus participant device; a bus receiving interface for receiving bus input data via a bus line; a bus transmitting interface for transmitting bus output data via the bus line; and a control unit for generating bus output data based on participant input data received via the bus participant interface, the bus transmission data including bus control data, and to transfer the bus output data to a further bus coupler by the bus transmitting interface. The control unit specifies a control signal based on the bus input data received by the bus receiving interface and performs a relaying of the bus input data to the further bus coupler based on the specified control signal.

Examples disclosed herein relate to managing a second ring link failure in a multi-ring Ethernet network. In an example, an inter-connection network node in a multi-ring Ethernet network comprising a major ring and a sub-ring may propagate a signal failure (SF) event, received in response to a second ring link failure in the major ring, to one or more nodes in the sub-ring. In response to receiving the SF event, a Ring Protection Link (RPL) on the sub-ring may be unlocked to allow network traffic through the RPL and avoid loop formation on the multi-ring Ethernet network. The sub-ring may be moved to the ring protection switching state; including performing a filtering database (FDB) flush at every node on the multi-ring Ethernet network whereby all MAC addresses and related port associations for traffic forwarding are cleared from the FDB.

Systems and methods for adding a new member to, and/or removing an existing member from, a ring of confederate hosts are provided. To add the new member, an invitation is sent from a current member, and instructions are sent to a sequentially prior confederate host with instructions to set a source address for further communications to an address of the new member. As data payloads are received, they are forwarded from the new member to the current member. To remove the existing member, instructions are sent to a sequentially prior confederate host with instructions to set a source address for further communications to an address of a sequentially next confederate host. As data payloads are received, they are forwarded from the sequentially prior confederate host to the sequentially next confederate hosts, cutting out the existing member.

A solution for obtaining an interior gateway protocol IGP domain through division in a network is provided. The network includes an access site link and a non-access site link. The access site link includes a link between access site devices and a link between an access site device and an aggregation site device. The non-access site link includes a link other than the access site link in the network. In this solution, a network device determines a changed access site subgraph based on a change of a network topology of the network, and obtains an IGP domain through division based on a link in the changed access site subgraph. The access site subgraph is one or more connection diagrams formed after the non-access site link is removed from the network topology. This solution can improve efficiency of IGP domain division.

An information obtaining unit (53) of a monitoring apparatus (40) obtains from each node, adjacent connection information (33) that is created in each node by an exchange of information and sharing of information between the nodes, and keeps the adjacent connection information (33) in a connection information file (38). The adjacent connection information (33) is information that indicates a connection relationship between the nodes. A topology comparison unit (54) of the monitoring apparatus (40) compares connection information included in the adjacent connection information (33) with topology definition information (59) retained in a topology definition file (58) beforehand, and determines whether or not a configuration of a ring network is according to design. The topology definition information (59) is information that defines the connection relationship between nodes in the ring network.

Embodiments of the present disclosure pertain to improved circuit and system architectures for identifying and managing operating statuses and faults in a system having multiple processing circuit chips. Each of the multiple processing circuit chips includes multiple signal rings, one to provide internal communications among circuitry within the circuit chip, and another with inter-chip communications circuitry to provide communications with neighboring circuit chips. One of the multiple processing circuit chips further includes external communications circuitry to provide communications with an external host.

A ring control data exchange system includes a first networking device in a first datacenter and a second networking device in a second datacenter and in a ring configuration with the first networking device. The second networking device identifies first control data packets, performs a hashing operation on the first control data packets to generate respective first control data hash values, and transmits the first control data packets to the first networking device. Subsequently, the second networking device identifies second control data packets, performs the hashing operation on the second control data packets to generate respective second control data hash values, determines that the respective second control data hash values match the respective first control data hash values and, in response, generates and transmits a first consolidated control data hash value to the first networking device while not transmitting the second control data packets to the first networking device.