Optimized class loading

A non-transitory computer-readable storage medium having computer-executable instructions stored thereupon which, when executed by one or more computers, cause the one or more computers to:

• build a first probabilistic data structure (PDS) for one or more classloaders;
• associate the first PDS with a first classloader of the one or more classloaders, the first PDS comprising false positive probability data indicating that the first classloader cannot load a first set of classes and that the first classloader has a first probability of being able to load a second set of classes;
• receive a request at the first classloader to load a requested class, the requested class included in the second set of classes;
• determine, based at least in part on the first PDS, a second probability that the first classloader can load the requested class;
• search a classpath associated with the first classloader for the requested class based at least in part on the second probability being above a predetermined threshold;
• determine, based at least in part on failing to identify the requested class within the classpath, that the first classloader cannot load the requested class;
• delegate the request to load the requested class to a second classloader, the second classloader associated with a second PDS;
• and
• load the requested class into a memory of the one or more computers.

An apparatus comprising:

• a processor;
• a memory;
• and
• a computer-readable storage medium having computer-executable instructions stored thereupon which, when executed by the processor, cause the apparatus to:
• access a probabilistic data structure for one or more classloaders;
• associate the probabilistic data structure with a first classloader of the one or more classloaders, the probabilistic data structure comprising false positive probability data indicating that the first classloader cannot load a first set of resources and that the first classloader has a first probability of being able to load a second set of resources;
• receive a request to load a requested resource into the memory;
• determine, based on the probabilistic data structure, a second probability that the first classloader can load the requested resource;
• search a classpath associated with the first classloader for the requested resource based at least in part on the second probability;
• determine, based at least in part on failing to identify the requested resource within the classpath or based at least in part on the second probability not meeting or exceeding a threshold, that the first classloader cannot load the requested resource;
• delegate the request to load the requested resource to a second classloader;
• and
• load the requested resource into the memory of the apparatus.