||Network resource reservation systems are being developed and deployed, driven by the demand and substantial benefits of providing performance predictability for modern distributed applications. However, existing systems suffer limitations: They either are inefficient in finding the optimal resource reservation, or cause private information (e.g., from the network infrastructure) to be exposed (e.g., to the user). In this paper, we design BoxOpt, a novel system that leverages efficient oracle construction techniques in optimization and learning theory to automatically, and swiftly learn the optimal resource reservations without exchanging any private information between the network and the user. In BoxOpt, we first model the simple reservation interface adopted in most reservation systems as a resource membership oracle. Second, we develop an efficient algorithm that constructs a resource separation oracle by a linear number of calls on resource membership oracle. Third, we develop a generic framework to construct a resource optimization oracle by iteratively calling the resource separation oracle, and then develop three novel, efficient algorithms under this generic framework, the best of which computes the optimal resource reservation by a linear number of calls on resource separation oracle. As such, BoxOpt can discover the optimal resource reservation with O(n2) calls on the resource membership oracle. We implement a prototype of BoxOpt with and demonstrate its efficiency and efficacy via extensive experiments using real network topology and a 7-day trace from a large operational federation network. Results show that (1) BoxOpt has a 100% correctness ratio by comparing with a state-of-the-art optimization solver, and (2) for 90% of requests, BoxOpt learns the optimal resource reservation within 10 seconds.