Abstract: A system and method are provided to customize allocation statistics. In one embodiment, a tag is associated to one or more threads to receive allocation statistic information relating to the one or more tagged threads, the one or more threads are tagged prior to performing an allocation statistic session. The allocation statistic session is performed via a plurality of threads at a first virtual machine at a server. A plurality of threads performing the allocation statistic session includes the one or more tagged threads is determined. The allocation statistic information relating to the one or more tagged threads is extracted, if the plurality of threads includes the one or more tagged threads.

Abstract: Methods and apparatus, including computer systems and program products, for sharing data objects in runtime systems. An identification of a first data object in a first runtime system is received. The first data object references zero or more referenced data objects. A shared closure of data objects, which consists of the first data object and a transitive closure of the referenced data objects, is identified, and a determination is made as to whether the shared closure of data objects is usable in a second runtime system. In some implementations, determining whether a shared closure is usable in a second runtime system includes determining whether each data object in the shared closure is serializable without execution of custom code, or determining whether the runtime class of each object instance in the shared closure is shareable. Using shared closures to share objects between runtime systems can provide isolation between user sessions.

Abstract: A system and method are provided to provide an enhanced memory error message. In one embodiment, a first message is associated to a memory error occurring at a virtual machine, the first message indicating the memory error has occurred. A second message is associated to the memory error when the memory error has occurred, the second message explaining the memory error. The memory error having the first message and the second message is thrown.

Abstract: Methods and apparatus, including computer systems and program products, for sharing classes and class loaders. One implementation provides a technique for performing a shared load procedure for a class, performing a post load procedure for the class, performing a shared link procedure for the class, performing a post link procedure for the class, and performing an initialization procedure for the class to reduce memory consumption and class loading time. Through performance of these procedures, a first portion of the class (which may include a class block and object) is loaded into a shared memory accessible by multiple runtime systems, and a second portion of the class (which may include a class static variable) as is stored in a local memory associated with a selected runtime system. A shared master copy of the second portion of the class is also stored in the shared memory.

Abstract: Methods and apparatus, including computer systems and program products to provide a Java virtual machine (VM) a extraction mechanism, which enables the Java VM to store a root set of information, which will also be referred to as exception data, in exception objects. The exception handler that catches the exception object can then call an extraction method to read the exception data from the exception object, extract additional information from the call stack using the exception data, and dump the exception data and extracted information into a file where it can be inspected later. The information dumped into the file can be used to facilitate development and debugging of Java-based software. Unlike other types of trace mechanisms utilized in Java systems, e.g., which may dump the entire core, the extraction mechanism enables a limited amount of information relating to specific exceptions to be stored for later inspection.

Abstract: Methods and apparatus, including computer systems and program products, for garbage collection of shared data entities. A computer program identifies a set of virtual machines (VMs), each operable to generate references to shared entities stored in shared memory. Each shared entity can include references to one or more of the other shared entities. The program waits for the VMs to independently identify the shared entities that they reference directly, and then initiates a shared garbage collection procedure to identify the shared entities that are not directly or indirectly referenced by any of the VMs and to reclaim the shared memory being used to store those shared entities. The VMs are prevented from generating new references to the shared entities solely during a portion of the shared garbage collection procedure. By minimizing the time during which access to the shared entities is restricted, performance and scalability can be increased.

Abstract: Methods and apparatus, including computer systems and program products, for sharing objects in runtime systems and for providing isolation between user sessions in a scalable manner. A user context corresponding to a user session is stored in a shared memory area. Upon receiving a request corresponding to the user session, a process is selected from a set of operating system processes, and a runtime system is selected from a set of runtime systems. The runtime system is bound to the process, and the user context is associated with the runtime system to process the request. In some implementations, sharing user contexts enables virtual machines to be used in a shared but isolated manner: Each user context can be exclusively associated with an available virtual machine from a fixed set of virtual machines, and each virtual machine can be exclusively executed in an available process from a fixed set of processes.

Abstract: Methods and apparatus, including computer systems and program products, for sharing data objects in runtime systems. An identification of a first data object in a first runtime system is received. The first data object references zero or more referenced data objects. A shared closure of data objects, which consists of the first data object and a transitive closure of the referenced data objects, is identified, and a determination is made as to whether the shared closure of data objects is usable in a second runtime system. In some implementations, determining whether a shared closure is usable in a second runtime system includes determining whether each data object in the shared closure is serializable without execution of custom code, or determining whether the runtime class of each object instance in the shared closure is shareable. Using shared closures to share objects between runtime systems can provide isolation between user sessions.

Abstract: Methods and apparatus, including computer systems and program products, for sharing objects in runtime systems. A set of objects is identified, each object having a reference to runtime metadata about the object in a first runtime system. The reference to the runtime metadata in each object is replaced with a proxy, and the set of objects is transmitted to a second runtime system, where the proxy in each object is replaced with a reference to runtime metadata about the object in the second runtime system. In some implementations, the runtime metadata can be installed in the second runtime system if it is not already available. The set of objects can include a shared closure of objects, which can be shared between runtime systems (including virtual machines) located on distinct physical machines. Sharing objects between runtime systems on multiple machines can provide isolation between user sessions in a scalable manner.