Abstract: A write protection cache (WPC) is introduced between a process and a disk cache on a solid state drive (SSD). The cache provides a holding area, and only data that has been seen more than once is committed to the disk cache. The WPC provides a layer of protection over the data on the disk by avoiding unnecessary writes to the drive. This protection is accomplished by only committing to disk the data or chunks that is seen more than once by a wide area network (WAN) process.

Abstract: A write protection cache (WPC) is introduced between a process and a disk cache on a solid state drive (SSD). The cache provides a holding area, and only data that has been seen more than once is committed to the disk cache. The WPC provides a layer of protection over the data on the disk by avoiding unnecessary writes to the drive. This protection is accomplished by only committing to disk the data or chunks that is seen more than once by a wide area network (WAN) process.

Abstract: After statically provisioned paths are created through an Ethernet network, trunks may be created to extend along those paths. The trunks extend between pairs of NNI interfaces on either end of the path, and Ethernet OAM may be run over the trunks to detect connectivity along the paths. Service instances and service groups may be provisioned over particular paths through the network. Each service instance/service group is associated with a trunk and the Ethernet OAM service on the trunk may be used to provide Ethernet connectivity OAM for the service instance/service group. Upon detection of a failure on the trunk, the service instances/service groups associated with the trunk may be caused to failover to an alternate path through the network. Thus, a single Ethernet OAM service instance may be used to provide connectivity verification so that connectivity detection does not need to be performed on each individual service instances/service group.

Abstract: After statically provisioned paths are created through an Ethernet network, trunks may be created to extend along those paths. The trunks extend between pairs of NNI interfaces on either end of the path, and Ethernet OAM may be run over the trunks to detect connectivity along the paths. Service instances and service groups may be provisioned over particular paths through the network. Each service instance/service group is associated with a trunk and the Ethernet OAM service on the trunk may be used to provide Ethernet connectivity OAM for the service instance/service group. Upon detection of a failure on the trunk, the service instances/service groups associated with the trunk may be caused to failover to an alternate path through the network. Thus, a single Ethernet OAM service instance may be used to provide connectivity verification so that connectivity detection does not need to be performed on each individual service instances/service group.

Abstract: After statically provisioned paths are created through an Ethernet network, trunks may be created to extend along those paths. The trunks extend between pairs of NNI interfaces on either end of the path, and Ethernet OAM may be run over the trunks to detect connectivity along the paths. Service instances and service groups may be provisioned over particular paths through the network. Each service instance/service group is associated with a trunk and the Ethernet OAM service on the trunk may be used to provide Ethernet connectivity OAM for the service instance/service group. Upon detection of a failure on the trunk, the service instances/service groups associated with the trunk may be caused to failover to an alternate path through the network. Thus, a single Ethernet OAM service instance may be used to provide connectivity verification so that connectivity detection does not need to be performed on each individual service instances/service group.

Abstract: An Ethernet switch for use in an Ethernet network comprises a set of ingress ports for receiving data frames and a set of egress ports. A memory is associated with each ingress port and stores forwarding information indicating one or more of the egress ports to which data frames received by that ingress port should be forwarded. A control or management interface receives information about a network connection established, or to be established, through the switch. A switch controller causes forwarding information to be stored in a memory associated with a first ingress port which the network connection will use, on the basis of the received information about a network connection. Forwarding information is not stored in a memory associated with a second of the set of ingress ports which the network connection will not use.

Abstract: An Ethernet switch for use in an Ethernet network comprises a set of ingress ports for receiving data frames and a set of egress ports. A memory is associated with each ingress port and stores forwarding information indicating one or more of the egress ports to which data frames received by that ingress port should be forwarded. A control or management interface receives information about a network connection established, or to be established, through the switch. A switch controller causes forwarding information to be stored in a memory associated with a first ingress port which the network connection will use, on the basis of the received information about a network connection. Forwarding information is not stored in a memory associated with a second of the set of ingress ports which the network connection will not use.

Abstract: A system for originating connectivity fault management (CFM) frames on non-CFM aware switches is disclosed. In the disclosed system, an OAM (Operations Administration and Management) proxy networking device connected to a core Ethernet network operates with one or more CPE (Customer Premises Equipment) devices to which it is communicably connected to support CFM out to the CPE devices without requiring that the CPE devices themselves implement CFM functionality. The OAM proxy networking device generates Connectivity Check (CC) frames, Loopback reply frames, and Linktrace reply frames for MEPs (Maintenance End Points) or MIPs (Maintenance Intermediate Points) contained in communication ports on the CPEs. These CFM frames generated by the OAM proxy networking device are then sent within messages from the OAM proxy networking device to the appropriate CPE.

Abstract: After statically provisioned paths are created through an Ethernet network, trunks may be created to extend along those paths. The trunks extend between pairs of NNI interfaces on either end of the path, and Ethernet OAM may be run over the trunks to detect connectivity along the paths. Service instances and service groups may be provisioned over particular paths through the network. Each service instance/service group is associated with a trunk and the Ethernet OAM service on the trunk may be used to provide Ethernet connectivity OAM for the service instance/service group. Upon detection of a failure on the trunk, the service instances/service groups associated with the trunk may be caused to failover to an alternate path through the network. Thus, a single Ethernet OAM service instance may be used to provide connectivity verification so that connectivity detection does not need to be performed on each individual service instances/service group.

Abstract: After statically provisioned paths are created through an Ethernet network, trunks may be created to extend along those paths. The trunks extend between pairs of NNI interfaces on either end of the path, and Ethernet OAM may be run over the trunks to detect connectivity along the paths. Service instances and service groups may be provisioned over particular paths through the network. Each service instance/service group is associated with a trunk and the Ethernet OAM service on the trunk may be used to provide Ethernet connectivity OAM for the service instance/service group. Upon detection of a failure on the trunk, the service instances/service groups associated with the trunk may be caused to failover to an alternate path through the network. Thus, a single Ethernet OAM service instance may be used to provide connectivity verification so that connectivity detection does not need to be performed on each individual service instances/service group.

Abstract: A network node includes a central processor card and a plurality of line cards. Each line card generates a maintenance association end point (MEP) entity that can respond to connectivity fault management (CFM) frames. The MEP entity on each line card periodically generates and transmits a multicast connectivity check message (CCM) to the other line cards in the network node. The CCM includes a card-information TLV and, optionally, a trunk-status TLV. Card-information TLVs include the slot number and card type of the transmitting line card. Trunk-status TLVs include the trunk state of each trunk supported by the transmitting line card. The line cards of the node consider a given line card to be down when three consecutive CCMs from that line card are missed. In response to recognizing a down line card, the other line cards can initiate an action, such as determine the trunks supported by the down line card and trigger a trunk switchover.

Abstract: A system for originating connectivity fault management (CFM) frames on non-CFM aware switches is disclosed. In the disclosed system, an OAM (Operations Administration and Management) proxy networking device connected to a core Ethernet network operates with one or more CPE (Customer Premises Equipment) devices to which it is communicably connected to support CFM out to the CPE devices without requiring that the CPE devices themselves implement CFM functionality. The OAM proxy networking device generates Connectivity Check (CC) frames, Loopback reply frames, and Linktrace reply frames for MEPs (Maintenance End Points) or MIPs (Maintenance Intermediate Points) contained in communication ports on the CPEs. These CFM frames generated by the OAM proxy networking device are then sent within messages from the OAM proxy networking device to the appropriate CPE.

Abstract: After statically provisioned paths are created through an Ethernet network, trunks may be created to extend along those paths. The trunks extend between pairs of NNI interfaces on either end of the path, and Ethernet OAM may be run over the trunks to detect connectivity along the paths. Service instances and service groups may be provisioned over particular paths through the network. Each service instance/service group is associated with a trunk and the Ethernet OAM service on the trunk may be used to provide Ethernet connectivity OAM for the service instance/service group. Upon detection of a failure on the trunk, the service instances/service groups associated with the trunk may be caused to failover to an alternate path through the network. Thus, a single Ethernet OAM service instance may be used to provide connectivity verification so that connectivity detection does not need to be performed on each individual service instances/service group.

Abstract: An Ethernet switch for use in an Ethernet network comprises a set of ingress ports for receiving data frames and a set of egress ports. A memory is associated with each ingress port and stores forwarding information indicating one or more of the egress ports to which data frames received by that ingress port should be forwarded. A control or management interface receives information about a network connection established, or to be established, through the switch. A switch controller causes forwarding information to be stored in a memory associated with a first ingress port which the network connection will use, on the basis of the received information about a network connection. Forwarding information is not stored in a memory associated with a second of the set of ingress ports which the network connection will not use.