Abstract: Various embodiments of apparatuses and methods for multi-cast, multiple unicast, and unicast distribution of messages with time synchronized delivery are described. In some embodiments, the disclosed system and methods include a reference timekeeper providing a reference clock to one or more host computing devices. The one or more host computing devices host compute instances, and also contain respective isolated timing hardware outside the control of the compute instances. The isolated timing hardware of the one or more host computing devices then receive respective packets, and obtain the same time to deliver the respective packets. Each isolated timing hardware provides either the packet, or information to access the packet, to its respective destination compute instance subsequent to determining that the same specified time to deliver the packet has occurred. Thus, the respective packets are delivered near simultaneously to the one or more destination compute instances.

Abstract: Systems and methods are provided for highly accurate synchronization of machine instances in a distributed, hosted computing environment to a reference timekeeper. In addition to a general communication network accessible to machine instances, the distributed environment includes a second network dedicated to carrying time information, such as a pulse-per-second (PPS) signal to isolated timing hardware within host computing devices. The isolated timing hardware can use the PPS signal, along with a reference time, to set a hardware clock. The isolated timing hardware can further provide an interface to machine instances that enables the instances to read the time of the hardware clock. This configuration enables many instances can share access to a single reference timekeeper, thus synchronizing those instances to a much higher accuracy than in traditional network-based time protocols.

Abstract: Systems and methods are provided for highly accurate synchronization of machine instances in a distributed, hosted computing environment to a reference timekeeper. In addition to a general communication network accessible to machine instances, the distributed environment includes a second network dedicated to carrying time information, such as a pulse-per-second (PPS) signal to isolated timing hardware within host computing devices. The isolated timing hardware can use the PPS signal, along with a reference time, to set a hardware clock. The isolated timing hardware can further provide an interface to machine instances that enables the instances to read the time of the hardware clock. This configuration enables many instances can share access to a single reference timekeeper, thus synchronizing those instances to a much higher accuracy than in traditional network-based time protocols.

Abstract: Various embodiments of apparatuses and methods for multi-cast, multiple unicast, and unicast distribution of messages with time synchronized delivery are described. In some embodiments, the disclosed system and methods include a reference timekeeper providing a reference clock to one or more host computing devices. The one or more host computing devices host compute instances, and also contain respective isolated timing hardware outside the control of the compute instances. The isolated timing hardware of the one or more host computing devices then receive respective packets, and obtain the same time to deliver the respective packets. Each isolated timing hardware provides either the packet, or information to access the packet, to its respective destination compute instance subsequent to determining that the same specified time to deliver the packet has occurred. Thus, the respective packets are delivered near simultaneously to the one or more destination compute instances.

Abstract: Various embodiments of apparatuses and methods for trusted and/or attested packet timestamping are described. In some embodiments, the disclosed system and methods include a reference timekeeper providing a reference clock to host computing devices. The host computing devices host compute instances using a first set of computing resources, and also contain isolated timing hardware utilizing a different set of computing resources. The isolated timing hardware sets a hardware clock based on a signal corresponding to the reference clock from the reference timekeeper. The isolated timing hardware then receives a packet from a particular compute instance, creates a timestamp for the packet based at least in part on the hardware clock, where the timestamp is outside the control of the compute instances, and sends the packet and the timestamp through a data network to transmit to a packet destination.

Abstract: Directional capacity of interfaces for networking devices are dynamically modified. Network traffic utilization of one direction of a network interface may be determined. A modification to a capacity of the networking device to process network traffic in the one direction of the network may be determined. The modification may then be applied to the networking device so that subsequent network traffic is processed according to the modified capacity in the one direction of the interface.

Abstract: Directional capacity of interfaces for networking devices are dynamically modified. Network traffic utilization of one direction of a network interface may be determined. A modification to a capacity of the networking device to process network traffic in the one direction of the network may be determined. The modification may then be applied to the networking device so that subsequent network traffic is processed according to the modified capacity in the one direction of the interface.

Abstract: The following description is directed to managing routing resources of a communication network. In one example, a method can include determining an amount of routing capacity from a network device and toward a routing prefix. An attribute associated with the routing prefix can be modified based on the amount of routing capacity toward the routing prefix. The modified attribute associated with the routing prefix can be announced to a peer of the network device.

Abstract: Optical transceivers include receiver-side and transmitter-side optical switches that selectively route optical signals to be transmitted to a wavelength division multiplexer or demultiplexed received signals to corresponding photodetectors. Other optical signals are transmitted or received via optical fibers and directed by the optical switches without wavelength multiplexing. In some examples, the switches are programmable and data routing can be altered as desired. In other examples, the switches are fixed, and data routing is altered by switch replacement.

Abstract: An emulation mode for network devices is described as a means of isolating a defective network device from real network traffic, while continuing to transmit faux traffic to the defective network device, wherein the faux traffic is intentionally dropped. The emulation mode allows the defective network device to be tested in an environment that appears real, without impacting user traffic. A management server can control one or more neighbor network devices and place them in the emulation mode as a way to isolate and test the defective network device. In the emulation mode, the neighbor network device monitors for packets that, based on previous routing history, would be sent to the defective network device and, instead, routes them to functioning network devices using a current FIB. A virtual RIB/FIB can also be used to route the same packets to the defective network device.

Abstract: Technologies are provided for ensuring that separate network paths within a network are used for network traffic between multiple source devices and a destination device or between pairs of source and destination devices. In some solutions, network path information is determined and sent to source devices that use the network path information to encapsulate network packets that then use separate network paths for at least a portion of their paths through the network. In some solutions, separate network paths are ensured using network path separation lists that are sent to network devices to make network traffic routing decisions. In some solutions, flow description information is calculated and sent to network devices to make network traffic routing decisions to keep network flows separate.

Abstract: Technologies are provided for ensuring that separate network paths within a network are used for network traffic between multiple source devices and a destination device or between pairs of source and destination devices. In some solutions, network path information is determined and sent to source devices that use the network path information to encapsulate network packets that then use separate network paths for at least a portion of their paths through the network. In some solutions, separate network paths are ensured using network path separation lists that are sent to network devices to make network traffic routing decisions. In some solutions, flow description information is calculated and sent to network devices to make network traffic routing decisions to keep network flows separate.

Abstract: Generally described, systems and methods are provided for detecting the impact of network failures. The system collects performance information from a plurality of nodes and links in a network, aggregates the collected performance information across paths in the network, processes the aggregated performance information for detecting failures on the paths, adjusts the set of performance information by removing the performance information for any nodes considered to be associated with performance information that is statistically different from performance information from other nodes at a given location or extrapolates the collected information to other paths, and determines the impact to customers of the network failures detected using the adjusted set of performance information.

Abstract: Technologies are provided for preventing transmission of cyclic redundancy check (CRC) errors, and/or other errors, within a computer network that uses cut-through forwarding of network packets. For example, counts of CRC errors can be maintained for network packets processed by network devices. When a CRC error condition is detected based on the counts of CRC errors, cut-through forwarding is disabled on at least a portion of the network (e.g., on one or more specific network ports and/or on one or more specific network devices). While cut-through forwarding is disabled, the affected portion of the network uses store-and-forward for processing network packets. When the CRC error condition no longer exists, then cut-through forwarding is re-enabled on the portion of the network.

Abstract: Encapsulated packets may be generated for different packets transmitted between a source instance and destination instance in a computer system. The source instance and destination instance may be implemented by different physical hosts linked by multiple network paths. Congestion of the multiple network paths may be determined and path-balancing polices may be implemented in response to the determined congestion. Each encapsulation packet comprises contents of a corresponding packet, and one or more data values selected in accordance with a path-balancing policy. The data values added to one encapsulation packet may differ from those added to another. Different network paths to the destination may be selected for different encapsulation packets of a given transmission based at least in part on the added data values.

Abstract: Methods and apparatus for Internet-scale routing using small-scale border routers and IP tunneling are described. Each border router is directly connected to a transit provider. Routing protocol peerings may be passed via the border routers through tunnels to a routing service; the routing service and the transit provider router(s) appear to be directly adjacent routing peers. The routing service receives routing data from the transit provider(s), maintains the routing data in a routing table, and processes the routing data in the routing table to select best paths. A mapping service may be informed, by the routing service, of a best exit point (or points) for each Internet prefix of each packet to be routed on the Internet. Outbound packets from devices on the network to the Internet, and inbound packets from the Internet to the network devices, may be encapsulated and passed through tunnels as directed by the mapping service.

Abstract: Encapsulated packets including sequencing information may be generated for packets to be transmitted between a source instance and destination instance. The source instance and destination instance may be implemented by different physical hosts linked by multiple network paths. Each encapsulated packet comprises contents of a corresponding packet, and one or more data values selected in accordance with whether sequencing is enable. Sequencing for network transmissions may be selective based at least in part one or more attributes of the source instance or destination instance.

Abstract: A plug may be coupled with a cable and include a fiber ferrule extending from the plug. A biasing mechanism may be arranged to bias a cover toward a covered configuration in which the fiber ferrule is situated within an internal volume defined by the cover and away from an uncovered configuration in which the fiber ferrule is situated at least partially outside of the internal volume defined by the cover. Bristles or other blockers may be positioned along or within a boundary of the internal volume of the cover, may be arranged to block particulate entry through the blockers into the internal volume of the cover in the covered configuration, and may be movable to permit passage of the fiber ferrule through the blockers in response to movement of the cover between the covered configuration and uncovered configuration.

Abstract: Generally described, systems and methods are provided for detecting the impact of network failures. The system collects performance information from a plurality of nodes and links in a network, aggregates the collected performance information across paths in the network, processes the aggregated performance information for detecting failures on the paths, adjusts the set of performance information by removing the performance information for any nodes considered to be associated with performance information that is statistically different from performance information from other nodes at a given location or extrapolates the collected information to other paths, and determines the impact to customers of the network failures detected using the adjusted set of performance information.

Abstract: A method for adjusting capacity in a multi-stage routing network includes monitoring a number of available connections between a router in a first stage of a multi-stage router network and one or more routers in a second stage of the multi-stage router network. Each of the stages of the multi-stage router network may include a plurality of routers. The method may also include detecting that the number of available connections falls below a threshold number. A notification can be sent to one or more routers in a third stage of the multi-stage router network that the router in the first stage is deprioritized. The one or more routers in the third stage can be operated so that communications to the first stage are routed to one or more other routers in the first stage.