End-to-end service level metric approximation

A method comprising:

• receiving end-to-end measurements of service performance for an application hosted on a network as a cloud service;
• identifying the application is experiencing degraded performance based on an anomaly in the end-to-end measurements;
• based on a plurality of applications experiencing degraded performance, wherein the application is included in the plurality of applications experiencing degraded performance, determining a service level metric by determining a count of the plurality of applications experiencing degraded performance exceeds a threshold number;
• determining the service level metric does not satisfy an expected service level associated with the application;
• and
• adjusting an allocation of network resources for the application to satisfy the expected service level.

A device comprising:

• a processing system;
• and
• memory comprising computer executable instructions that, when executed, perform operations comprising:
• receiving end-to-end measurements of service performance for an application hosted on a network;
• identifying the application is experiencing degraded performance based on an anomaly in the end-to-end measurements;
• determining a service level metric for the application by determining a count of applications experiencing degraded performance exceeds a threshold number;
• determining the service level metric does not satisfy an expected service level associated with the application;
• and
• in response to determining the service level metric does not satisfy the expected service level, adjusting an allocation of network resources for the application.