Network pathway diagnosis
Techniques are disclosed for network pathway diagnosis. The system accesses sets of training data. The sets of training data define associations between issues of computer networks and network pathways of the computer networks. The training data is used to train a machine learning model to select for diagnosis network pathways in computer networks. An issue is detected in a computer network that incorporates a cluster of computing nodes configured for executing containerized applications. The trained machine learning model is applied to select a target network pathway in the computer network for diagnosis based on a target set of characteristics that are associated with the issue. Diagnosing the target network pathway reveals that the target network pathway is dysfunctional, why the target network pathway is dysfunctional, and/or actions that potentially resolve the dysfunction of the target network pathway.
1 . One or more non-transitory computer-readable media comprising instructions that, when executed by one or more hardware processors, cause performance of operations comprising:
accessing sets of training data, a first set of training data of the sets of training data defining an association between (a) a first issue relating to a first cluster of computing nodes configured for executing a first set of containerized applications and (b) one or more network pathways of a first computer network, wherein the first computer network comprises the first cluster of computing nodes;
training a machine learning model to select network pathways in computer networks for diagnosis based on the sets of training data;
detecting a second issue, the second issue relating to a second cluster of computing nodes configured for executing a second set of containerized applications, the second cluster of computing nodes comprised within a second computer network;
accessing a target set of characteristics associated with the second issue;
applying the machine learning model to the target set of characteristics to select a target network pathway in the second computer network for diagnosis; and
based, at least in part, on applying the machine learning model to select the target network pathway, diagnosing the target network pathway, wherein diagnosing the target network pathway comprises:
executing one or more diagnostics corresponding to the target network pathway, wherein executing a first diagnostic of the one or more diagnostics comprises evaluating connectivity of the target network pathway by analyzing a configuration of the second computer network.
2 . The one or more non-transitory computer-readable media of claim 1 :
wherein the second issue is resolved based at least in part on the diagnosis of the target network pathway.
3 . The one or more non-transitory computer-readable media of claim 2 , wherein evaluating the connectivity of the target network pathway by analyzing the configuration of the second computer network comprises:
accessing routing information and security rules of the second computer network; and
predicting a connectivity status of the target network pathway based, at least in part, on the routing information and/or the security rules.
4 . The one or more non-transitory computer-readable media of claim 2 , wherein the operations further comprise:
prior to executing the one or more diagnostics corresponding to the target network pathway:
verifying that one or more network policies are applied to the second computer network, wherein the one or more network policies enable the one or more diagnostics corresponding to the target network pathway to be successfully executed; and
subsequent to executing the one or more diagnostics corresponding to the target network pathway:
presenting results of the one or more diagnostics on a graphical user interface (GUI).
5 . The one or more non-transitory computer-readable media of claim 2 , wherein the operations further comprise:
determining a second root cause of the second issue based, at least in part, on the one or more diagnostics corresponding to the target network pathway;
determining one or more actions that may resolve the second issue;
prompting a generative AI model to output a description of the one or more actions that may resolve the second issue; and
presenting the description of the one or more actions that may resolve the second issue on a GUI.
6 . The one or more non-transitory computer-readable media of claim 5 , wherein the target network pathway connects at least a first component comprised within the second cluster of computing nodes to a second component comprised within the second computer network, wherein the second root cause of the second issue is a misconfiguration of the second cluster of computing nodes, and wherein the one or more actions comprise reconfiguring a security configuration and/or a routing configuration of the second cluster of computing nodes.
7 . The one or more non-transitory computer-readable media of claim 1 , wherein the first cluster of computing nodes is a Kubernetes cluster.
8 . The one or more non-transitory computer-readable media of claim 1 , wherein detecting the second issue comprises receiving natural language user input describing the second issue, and wherein accessing a target set of characteristics associated with the second issue comprises applying natural language processing to the natural language user input to generate a second feature set.
9 . The one or more non-transitory computer-readable media of claim 8 , wherein the first set of training data comprises (a) a first feature set generated by applying natural language processing to a service ticket comprising text data describing the first issue and (b) the one or more network pathways of the first computer network that are associated with the first issue, and wherein the one or more network pathways associated with the first issue are determined to correspond to a first root cause of the first issue.
10 . The one or more non-transitory computer-readable media of claim 1 , wherein the target set of characteristics associated with the second issue comprises at least one of (a) a type of the second issue, (b) a location of the second issue, (c) a timing of the second issue, (d) a sequence of events that are temporally, causally, and/or topologically related to the second issue, (e) a topology of the second computer network, (f) network policies of the second computer network, (g) a capacity of the second computer network, (h) a network load of the second computer network, (i) historical activity of a user of the second cluster of computing nodes, and/or (j) a user characteristic of the user of the second cluster of computing nodes.
11 . The one or more non-transitory computer-readable media of claim 1 , wherein the operations further comprise:
accessing feedback pertaining to the selecting of the target network pathway by the machine learning model,
wherein the feedback comprises at least one of (a) user input, (b) results of the one or more diagnostics corresponding to the target network pathway, (c) a predicted connectivity status of the target network pathway, and/or (d) an indication of whether the second issue is resolved based, at least in part, on diagnosing the target network pathway; and
further training the machine learning model based at least in part on the feedback.
12 . A method, comprising:
accessing sets of training data, a first set of training data of the sets of training data defining an association between (a) a first issue relating to a first cluster of computing nodes configured for executing a first set of containerized applications and (b) one or more network pathways of a first computer network, wherein the first computer network comprises the first cluster of computing nodes;
training a machine learning model to select network pathways in computer networks for diagnosis based on the sets of training data;
detecting a second issue, the second issue relating to a second cluster of computing nodes configured for executing a second set of containerized applications, the second cluster of computing nodes comprised within a second computer network;
accessing a target set of characteristics associated with the second issue;
applying the machine learning model to the target set of characteristics to select a target network pathway in the second computer network for diagnosis; and
based, at least in part, on applying the machine learning model to select the target network pathway, diagnosing the target network pathway, wherein diagnosing the target network pathway comprises:
executing one or more diagnostics corresponding to the target network pathway, wherein executing a first diagnostic of the one or more diagnostics comprises evaluating connectivity of the target network pathway by analyzing a configuration of the second computer network,
wherein the method is performed by at least one device including a hardware processor.
13 . The method of claim 12 :
wherein the second issue is resolved based at least in part on the diagnosis of the target network pathway.
14 . The method of claim 13 , wherein evaluating the connectivity of the target network pathway by analyzing the configuration of the second computer network comprises:
accessing routing information and security rules of the second computer network; and
predicting a connectivity status of the target network pathway based, at least in part, on the routing information and/or the security rules.
15 . The method of claim 12 , wherein the first cluster of computing nodes is a Kubernetes cluster, wherein detecting the second issue comprises receiving natural language user input describing the second issue, and wherein accessing a target set of characteristics associated with the second issue comprises applying natural language processing to the natural language user input to generate a second feature set.
16 . The method of claim 15 , wherein the first set of training data comprises (a) a first feature set generated by applying natural language processing to a service ticket comprising text data describing the first issue and (b) the one or more network pathways of the first computer network that are associated with the first issue, and wherein the one or more network pathways associated with the first issue are determined to correspond to a first root cause of the first issue.
17 . The method of claim 12 , wherein the target set of characteristics associated with the second issue comprises at least one of (a) a type of the second issue, (b) a location of the second issue, (c) a timing of the second issue, (d) a sequence of events that are temporally, causally, and/or topologically related to the second issue, (e) a topology of the second computer network, (f) network policies of the second computer network, (g) a capacity of the second computer network, (h) a network load of the second computer network, (i) historical activity of a user of the second cluster of computing nodes, and/or (j) a user characteristic of the user of the second cluster of computing nodes.
18 . A system, comprising:
at least one device including a hardware processor;
the system being configured to perform operations comprising:
accessing sets of training data, a first set of training data of the sets of training data defining an association between (a) a first issue relating to a first cluster of computing nodes configured for executing a first set of containerized applications and (b) one or more network pathways of a first computer network, wherein the first computer network comprises the first cluster of computing nodes;
training a machine learning model to select network pathways in computer networks for diagnosis based on the sets of training data;
detecting a second issue, the second issue relating to a second cluster of computing nodes configured for executing a second set of containerized applications, the second cluster of computing nodes comprised within a second computer network;
accessing a target set of characteristics associated with the second issue;
applying the machine learning model to the target set of characteristics to select a target network pathway in the second computer network for diagnosis; and
based, at least in part, on applying the machine learning model to select the target network pathway, diagnosing the target network pathway, wherein diagnosing the target network pathway comprises:
executing one or more diagnostics corresponding to the target network pathway, wherein executing a first diagnostic of the one or more diagnostics comprises evaluating connectivity of the target network pathway by analyzing a configuration of the second computer network.