Network monitoring combining client-based and network-based signals

A method, comprising:

• receiving a request to monitor a network within a compute service provider, the request including two endpoints within the network and a signal of client-based monitoring data for the network, the client-based monitoring data including performance signals from one or more client-based monitors indicating one or more parameters of client traffic or events experienced by client devices attempting to access a destination in the network;
• in response to detecting a performance degradation in the network based on the client-based monitoring data, comparing the client-based monitoring data received via the signal with network-based monitoring data for the network to determine if one or more events indicated by the network-based monitoring data are associated with or are a root cause for the performance degradation detected based on the client-based monitoring data;
• transmitting a report indicating a result of the comparison, including an indication of whether the network-based monitoring data correlated with the client-based monitoring data confirms the performance degradation is related to an issue in the network;
• and
• responsive to confirming the performance degradation is related to the issue in the network, performing a mitigation for the issue.

One or more computer-readable storage media comprising computer-executable instructions that, when executed, cause a computing system to perform a method comprising:

• receiving a request from a client to perform network monitoring using a compute service provider, the request including a pair of endpoints in a network and a signal of client-based monitoring data captured for traffic sent between the pair of endpoints from a perspective of a client accessing the network;
• computing routes through the network between the pair of endpoints and aggregating network-based monitoring data for traffic sent along the computed routes;
• identifying one or more events on one or more paths between the pair of endpoints using the network-based monitoring data;
• in response to a first detection that performance signals indicated by the client-based monitoring data correspond to a network issue, performing a first comparison correlating the client-based monitoring data received via the signal with at least one of the one or more events indicated by corresponding network-based monitoring data to confirm whether the performance signals are caused by the network issue or by an issue outside of the network;
• in response to a second detection, subsequent to the first detection, that the performance signals indicated by the client-based monitoring data correspond to the network issue, injecting probes into the network to generate targeted network-based monitoring data for a reduced subset of the computed routes through the network between the pair of endpoints and performing a second comparison correlating the targeted network-based monitoring data with the client-based monitoring data received via the signal to confirm whether the performance signals are caused by the network issue;
• and
• transmitting a report to the client indicating whether the performance signals are confirmed to be caused by the network issue based on the first and second comparisons.