Abstract: A network administration device may include one or more processors to receive operational information regarding a plurality of network devices; receive flow information relating to at least one traffic flow; input the flow information to a model, where the model is generated based on a machine learning technique, and where the model is configured to identify predicted performance information of one or more network devices with regard to the at least one traffic flow based on the operational information; determine path information for the at least one traffic flow with regard to the one or more network devices based on the predicted performance information; and/or configure the one or more network devices to implement the path information for the traffic flow.

Abstract: In some examples, a switching system includes a plurality of fabric endpoints and a multi-stage switching fabric having a plurality of fabric planes each having a plurality of stages to switch data units between any of the plurality of fabric endpoints. A fabric endpoint of the fabric endpoints is configured to send, to a switch of a first one of the stages and within a first fabric plane of the plurality of fabric planes, a self-ping message destined for the fabric endpoint. The fabric endpoint is configured to send, in response to determining the fabric endpoint has not received the self-ping message after a predetermined time, an indication of a connectivity fault for the first fabric plane.

Abstract: A device may receive a trained data model that has been trained using historical link quality information associated with a set of links. The device may determine, after receiving the trained data model, link quality information associated with a link that is actively supporting traffic. The device may classify the link by using the link quality information as input for the data model. The data model may classify the link into a class of a set of classes associated with measuring link quality. The device may determine an actual quality level of the link. The device may selectively update the class of the link after determining the actual link quality of the link. The device may perform one or more actions associated with improving link quality based on classifying the link and/or selectively updating the class of the link.

Abstract: A network device may include a packet generator device implemented in hardware. The packet generator device may include a control component, a payload generation component, and an interface element to receive test packet generation information. The test packet generation information may include one or more control inputs, header data that is to be included in one or more test packets, and information regarding a data pattern that is to be included in payload data of the one or more test packets. The one or more control inputs, when provided to the control component, may cause the control component to control the payload generation component to generate the one or more test packets based on the header data and the information regarding the data pattern.

Abstract: In some examples, a switching system includes a plurality of fabric endpoints and a multi-stage switching fabric having a plurality of fabric planes each having a plurality of stages to switch data units between any of the plurality of fabric endpoints. A fabric endpoint of the fabric endpoints is configured to send, to a switch of a first one of the stages and within a first fabric plane of the plurality of fabric planes, a self-ping message destined for the fabric endpoint. The fabric endpoint is configured to send, in response to determining the fabric endpoint has not received the self-ping message after a predetermined time, an indication of a connectivity fault for the first fabric plane.

Abstract: A device may receive a trained data model that has been trained using historical link quality information associated with a set of links. The device may determine, after receiving the trained data model, link quality information associated with a link that is actively supporting traffic. The device may classify the link by using the link quality information as input for the data model. The data model may classify the link into a class of a set of classes associated with measuring link quality. The device may determine an actual quality level of the link. The device may selectively update the class of the link after determining the actual link quality of the link. The device may perform one or more actions associated with improving link quality based on classifying the link and/or selectively updating the class of the link.

Abstract: A network device may include a packet generator device implemented in hardware. The packet generator device may include a control component, a payload generation component, and an interface element to receive test packet generation information. The test packet generation information may include one or more control inputs, header data that is to be included in one or more test packets, and information regarding a data pattern that is to be included in payload data of the one or more test packets. The one or more control inputs, when provided to the control component, may cause the control component to control the payload generation component to generate the one or more test packets based on the header data and the information regarding the data pattern.

Abstract: In some examples, a switching system includes a plurality of fabric endpoints and a multi-stage switching fabric having a plurality of fabric planes each having a plurality of stages to switch data units between any of the plurality of fabric endpoints. A fabric endpoint of the fabric endpoints is configured to send, to a switch of a first one of the stages and within a first fabric plane of the plurality of fabric planes, a self-ping message destined for the fabric endpoint. The fabric endpoint is configured to send, in response to determining the fabric endpoint has not received the self-ping message after a predetermined time, an indication of a connectivity fault for the first fabric plane.

Abstract: A device may receive a trained data model that has been trained using historical link quality information associated with a set of links. The device may determine, after receiving the trained data model, link quality information associated with a link that is actively supporting traffic. The device may classify the link by using the link quality information as input for the data model. The data model may classify the link into a class of a set of classes associated with measuring link quality. The device may determine an actual quality level of the link. The device may selectively update the class of the link after determining the actual link quality of the link. The device may perform one or more actions associated with improving link quality based on classifying the link and/or selectively updating the class of the link.

Abstract: A system may comprise a first device and a second device associated with a Clos architecture. The first device may include a first crossbar that comprises a first component, a second component, and a third component. The second device may include a second crossbar that comprises a fourth component, a fifth component, and a sixth component. The first component may connect to the second component and the fifth component. The second component may connect to the first component, the third component, the fourth component, and the sixth component. The third component may connect to the second component and the fifth component. The fourth component may connect to the second component and the fifth component. The fifth component may connect to the first component, the third component, the fourth component, and the sixth component. The sixth component may connect to the second component and the fifth component.

Abstract: A device may receive a trained data model that has been trained using historical link quality information associated with a set of links. The device may determine, after receiving the trained data model, link quality information associated with a link that is actively supporting traffic. The device may classify the link by using the link quality information as input for the data model. The data model may classify the link into a class of a set of classes associated with measuring link quality. The device may determine an actual quality level of the link. The device may selectively update the class of the link after determining the actual link quality of the link. The device may perform one or more actions associated with improving link quality based on classifying the link and/or selectively updating the class of the link.

Abstract: A network administration device may include one or more processors to receive operational information regarding a plurality of network devices; receive flow information relating to at least one traffic flow; input the flow information to a model, where the model is generated based on a machine learning technique, and where the model is configured to identify predicted performance information of one or more network devices with regard to the at least one traffic flow based on the operational information; determine path information for the at least one traffic flow with regard to the one or more network devices based on the predicted performance information; and/or configure the one or more network devices to implement the path information for the traffic flow.

Abstract: In some examples, a switching system includes a plurality of fabric endpoints and a multi-stage switching fabric having a plurality of fabric planes each having a plurality of stages to switch data units between any of the plurality of fabric endpoints. A fabric endpoint of the fabric endpoints is configured to send, to a switch of a first one of the stages and within a first fabric plane of the plurality of fabric planes, a self-ping message destined for the fabric endpoint. The fabric endpoint is configured to send, in response to determining the fabric endpoint has not received the self-ping message after a predetermined time, an indication of a connectivity fault for the first fabric plane.

Abstract: In some examples, a switching system includes a plurality of fabric endpoints and a multi-stage switching fabric having a plurality of fabric planes each having a plurality of stages to switch data units between any of the plurality of fabric endpoints. A fabric endpoint of the fabric endpoints is configured to send, to a switch of a first one of the stages and within a first fabric plane of the plurality of fabric planes, a self-ping message destined for the fabric endpoint. The fabric endpoint is configured to send, in response to determining the fabric endpoint has not received the self-ping message after a predetermined time, an indication of a connectivity fault for the first fabric plane.

Abstract: An example network device includes a set of physical network interfaces and a control unit that executes a routing protocol and a traffic impact prediction module. The traffic impact prediction module determines, prior to occurrence of a topology-changing device fault, that one or more operating characteristics of the network device are indicative of a possible fault, wherein the network device is one of a plurality of network devices in a network, determines a probability of traffic loss associated with the possible fault, and determines an adjusted routing metric for routes impacted by the possible fault based at least in part on the probability of the traffic loss. The routing protocol sends, via at least one of the set of physical network interfaces, one or more interior gateway protocol update messages specifying the adjusted routing metric to at least one other network device in the network.

Abstract: A system may comprise a first device and a second device associated with a Clos architecture. The first device may include a first crossbar that comprises a first component, a second component, and a third component. The second device may include a second crossbar that comprises a fourth component, a fifth component, and a sixth component. The first component may connect to the second component and the fifth component. The second component may connect to the first component, the third component, the fourth component, and the sixth component. The third component may connect to the second component and the fifth component. The fourth component may connect to the second component and the fifth component. The fifth component may connect to the first component, the third component, the fourth component, and the sixth component. The sixth component may connect to the second component and the fifth component.

Abstract: In some examples, a switching system includes a plurality of fabric endpoints and a multi-stage switching fabric having a plurality of fabric planes each having a plurality of stages to switch data units between any of the plurality of fabric endpoints. A fabric endpoint of the fabric endpoints is configured to send, to a switch of a first one of the stages and within a first fabric plane of the plurality of fabric planes, a self-ping message destined for the fabric endpoint. The fabric endpoint is configured to send, in response to determining the fabric endpoint has not received the self-ping message after a predetermined time, an indication of a connectivity fault for the first fabric plane.

Abstract: A system may comprise a first device and a second device associated with a Clos architecture. The first device may include a first crossbar that comprises a first component, a second component, and a third component. The second device may include a second crossbar that comprises a fourth component, a fifth component, and a sixth component. The first component may connect to the second component and the fifth component. The second component may connect to the first component, the third component, the fourth component, and the sixth component. The third component may connect to the second component and the fifth component. The fourth component may connect to the second component and the fifth component. The fifth component may connect to the first component, the third component, the fourth component, and the sixth component. The sixth component may connect to the second component and the fifth component.