# US 12,015,534 B2 · Method and system for congestion detection and validation in a network > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,015,534 B2 - Document ID: US12015534B2 - Status: Granted - Kind: B2 - Application number: 17/156,194 - Filed: 2021-01-22 - Published: 2022-07-28 - Granted: 2024-06-18 - Earliest claimed: 2021-01-22 - Assignees: Arista Networks, Inc. - Applicants: Arista Networks, Inc. - Inventors: Gagandeep S. Arneja - Web page: https://patentgrep.com/patent/US12015534B2 ## CPC - H04L 43/0817: by checking functioning - H04L 41/0627: {by acting on the notification or alarm source} - H04L 43/067: using time frame reporting - H04L 43/0852: Delays - H04L 43/0876: Network utilisation, e.g. volume of load or congestion level - H04L 47/11: Identifying congestion - H04L 47/12: Avoiding congestion; Recovering from congestion - H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M) ## Abstract A method for managing a network by a network monitoring system, wherein the network comprises a plurality of network devices, the method comprising receiving, by the network monitoring system, in-band network telemetry (INT) data from a network device of the plurality of network devices, updating a latency model of the network using the INT data to obtain an updated latency model, identifying a congestion point in the network using the updated latency model and at least a portion of the INT data, validating the congestion point; and initiating a remediation action based on the validation. ## Claims ### Claim 1 A method for managing a network by a network monitoring system, wherein the network comprises a plurality of network devices, the method comprising: receiving, by the network monitoring system, in-band network telemetry (INT) data from a network device of the plurality of network devices; dynamically updating a latency model of the network using the INT data to obtain an updated latency model, the latency model comprising latency values for egress queues on the plurality of network devices, the latency values determined based on latency measurements for the egress queues determined from the INT data; identifying a congestion point associated with a network device in the network using the updated latency model and at least a portion of the INT data; validating the congestion point, wherein validating the congestion point comprises determining whether a queue depth report associated with the congestion point has been received within a validation window; and initiating a remediation action based on the validation. ### Claim 2 (depends on 1) The method of claim 1, wherein the congestion point is validated using queue depth monitoring at the network device associated with the congestion point. ### Claim 3 (depends on 1) The method of claim 1, wherein a latency value is an exponentially weighted moving average of latency measurements for at least one queue, wherein second INT data is used to derive a latency measurement of the latency measurements for the at least one queue, wherein the second INT data is received from the network device prior to the INT data. ### Claim 4 (depends on 1) The method of claim 1, wherein the congestion point is identified using the latency value in the latency model and the latency measurement derived from the INT data. ### Claim 5 (depends on 4) The method of claim 4, wherein the congestion point is identified when the latency measurement exceeds the latency value. ### Claim 6 (depends on 4) The method of claim 4, wherein the congestion point is identified when the latency measurement deviates from the latency value by more than a range. ### Claim 7 (depends on 4) The method of claim 4, wherein the congestion point is associated with an egress queue on one of the plurality of network devices. ### Claim 8 (depends on 1) The method of claim 1, wherein the validation window comprises a period of time that includes a time at which the congestion point was identified. ### Claim 9 (depends on 8) The method of claim 8, wherein the validation window comprises a second time that is later than the time at which the congestion point was identified, wherein the queue depth report associated with the congestion point was received at the second time. ### Claim 10 (depends on 1) The method of claim 1, wherein the remediation action comprises issuing a congestion notification. ### Claim 11 (depends on 10) The method of claim 10, further comprising: obtaining flow tracking information for at least one network device in the network, wherein the congestion notification specifies at least one flow, wherein the flow tracking information specifies the at least one flow. ### Claim 12 (depends on 1) The method of claim 1, wherein the remediation action comprises initiating a corrective action on the network. ### Claim 13 (depends on 12) The method of claim 12, wherein the corrective action comprises modifying an operation of at least one network device in the network. ### Claim 14 A method for managing a network, the method comprising: receiving, by a network monitoring system, in-band network telemetry (INT) data from the network; identifying a congestion point associated with a network device in the network using a dynamically updated latency model and at least a portion of the INT data, the latency model comprising latency values for egress queues on the plurality of network devices, the latency values determined based on latency measurements for the egress queues determined from the INT data; validating the congestion point using a queue depth report; obtaining flow tracking information, wherein the flow information specifies the at least one flow associated with the congestion point; and issuing, in response to the identifying, a congestion notification specifying the at least one flow. ### Claim 15 (depends on 14) The method of claim 14, further comprising: making a determination that a queue depth report associated with the congestion point is not received during a validation window, wherein the determination does not prevent the congestion notification from being issued, and wherein the congestion notification specifies that the congestion point was not validated. ### Claim 16 (depends on 14) The method of claim 14, wherein the congestion point is an egress queue in the network. ### Claim 17 (depends on 14) The method of claim 14, wherein the congestion point is identified when the latency measurement deviates from the latency value. ### Claim 18 A network monitoring system operatively connected to a plurality of network devices in a network, comprising: a processor, memory comprising instructions, which when executed by the processor, perform a method, the method comprising: receiving in-band network telemetry (INT) data from the network; identifying a congestion point associated with a network device in the network using a dynamically updated latency model and at least a portion of the INT data, the latency model comprising latency values for egress queues on the plurality of network devices, the latency values determined based on latency measurements for the egress queues determined from the INT data; validating the congestion point using a queue depth report obtained from the network; obtaining flow tracking information for at least one network device in the network the flow tracking information specifying at least on flow associated with the congestion point; and initiating a remediation action based on the validation, wherein the remediation action comprises issuing a congestion notification specifying the at least one flow.