A system includes a first and at least one second processing circuit, a configuration engine, and a switch. The configuration engine stores a virtual link configuration for a plurality of virtual links, which indicates a priority and a predetermined network pathway for communicating data packets from the first processing circuit to the at least one second processing circuit. The configuration engine automatically assigns priority to a first virtual link of the plurality of virtual links based on at least one of latency or jitter. The switch receives a first data packet from the first processing circuit. A first virtual link identifier is extracted from the first data packet. A first priority and a first predetermined network pathway corresponding to the first virtual link identifier from the virtual link configuration are retrieved. The first data packet is transmitted along the first predetermined network pathway based on the first priority.

Systems including at least one processor and a memory storing instructions that, when executed by the at least one processor, result in the system collecting real-time telemetry measurements for packets received at each hop of an overlay network, and the system injecting the measurements into a variable-length trailers of the packets.

A method for autonomous data routing in a distributed network includes installing containerized applications at a plurality of nodes including a first node, a second node, and a relay node in a computer network, automatically measuring one-way latencies between the plurality of nodes responsive to instructions of the containerized applications, automatically selecting, responsive to the containerized applications, a relayed data routing path from the first node to the second node via the relay node at least in part based on the one-way latencies between nodes in the computer network, automatically transferring data from the first node to the second node along the relayed data routing path responsive to instructions of the containerized applications, and in response to the data transfer, automatically transferring a payment between digital wallets under the control of the containerized applications.

A method for autonomous data routing in a distributed network includes installing containerized applications at a plurality of nodes including a first node, a second node, and a relay node in a computer network, automatically measuring one-way latencies between the plurality of nodes responsive to instructions of the containerized applications, automatically selecting, responsive to the containerized applications, a relayed data routing path from the first node to the second node via the relay node at least in part based on the one-way latencies between nodes in the computer network, automatically transferring data from the first node to the second node along the relayed data routing path responsive to instructions of the containerized applications, and in response to the data transfer, automatically transferring a payment between digital wallets under the control of the containerized applications.

Various approaches for allocating resources to an application having multiple application components, with at least one executing one or more functions, in a serverless service architecture include identifying multiple routing paths, each routing path being associated with a same function service provided by one or more containers or serverless execution entities; determining traffic information on each routing path and/or a cost, a response time and/or a capacity associated with the container or serverless execution entity on each routing path; selecting one of the routing paths and its associated container or serverless execution entity; and causing a computational user of the application to access the container or serverless execution entity on the selected routing path and executing the function(s) thereon.

Various approaches for allocating resources to an application having multiple application components, with at least one executing one or more functions, in a serverless service architecture include identifying multiple routing paths, each routing path being associated with a same function service provided by one or more containers or serverless execution entities; determining traffic information on each routing path and/or a cost, a response time and/or a capacity associated with the container or serverless execution entity on each routing path; selecting one of the routing paths and its associated container or serverless execution entity; and causing a computational user of the application to access the container or serverless execution entity on the selected routing path and executing the function(s) thereon.

A path establishment method and a controller are disclosed. The method includes: when detecting a path establishment request for establishing P2MP TE, computing a P2MP TE path by using head node information and tail node information included in the path establishment request; identifying a target branch node in the P2MP TE path, and obtaining a label of the target branch node; and when a third node corresponding to the head node information and the target branch node are not a same node, sending first information to the third node, and sending second information to the target branch node, where the second information is used to instruct the target branch node to generate a multicast forwarding entry. Embodiments of this application can reduce complexity of establishing the P2MP TE path.

A path establishment method and a controller are disclosed. The method includes: when detecting a path establishment request for establishing P2MP TE, computing a P2MP TE path by using head node information and tail node information included in the path establishment request; identifying a target branch node in the P2MP TE path, and obtaining a label of the target branch node; and when a third node corresponding to the head node information and the target branch node are not a same node, sending first information to the third node, and sending second information to the target branch node, where the second information is used to instruct the target branch node to generate a multicast forwarding entry. Embodiments of this application can reduce complexity of establishing the P2MP TE path.

A method of managing multipath traffic switching in a network includes a plurality of shared paths, including selecting a first path from among the plurality of shared paths; setting an initial timer to an initial timer value; sending network traffic over the first path; based on determining that at least one path characteristic of the first path is unsatisfactory and that the initial timer has not expired, generating an updated timer value by increasing the initial timer value; and based on determining that the at least one path characteristic of the first path is unsatisfactory and that the initial timer has expired, selecting a second path, sending the network traffic over the second path, and setting an updated timer to the updated timer value.

A method of managing multipath traffic switching in a network includes a plurality of shared paths, including selecting a first path from among the plurality of shared paths; setting an initial timer to an initial timer value; sending network traffic over the first path; based on determining that at least one path characteristic of the first path is unsatisfactory and that the initial timer has not expired, generating an updated timer value by increasing the initial timer value; and based on determining that the at least one path characteristic of the first path is unsatisfactory and that the initial timer has expired, selecting a second path, sending the network traffic over the second path, and setting an updated timer to the updated timer value.