Abstract: Various techniques are described for updating a request during propagation among services to include different priority values. Each service may determine its own priority for a request and add that priority to the request's overall priority value before propagating the request to the next service. When an overloaded service receives multiple requests, it may order the requests into a queue based on the different priority values. This may allow multiple different services to affect the priority/importance of any given request. A service may use the different priorities in order to determine the lowest priority requests that should be dropped when a queue needs to make room for a higher priority request.