Research Of Cold Object In Java Virtual Machine

Java as a technology that has the most user and be used widely,the main reason is a advantage that greatly reduce the develop cost: it is running in JVM(java virtual machine,and JVM hide the complexity of basement hardware and the difference of the OS(operating system),that can create an unitive environment at different physic machine,achieve the goal that “write once,run anywhere”.It makes programmers can concentrate on the business logic,the work will more efficiently.the same as to improve work efficiency,JVM provide strong memory management function,because include the garbage collection,java programmers not need to care garbage collection any more like c/c++ programmers.But it is because that JVM plays a important role,so that JVM optimize become a important issue.Many Java applications instantiate objects within the Java heap that are persistent but seldom if ever referenced by the application.Examples include strings,such as error messages,and collections of value objects that are pre-loaded for fast access but they may include objects that are seldom referenced.The cold object is identified and estimated by establishing a activity map to trace the activity of the object in heap and compare to the reset parameter Tcold.Build the visited model is more fit to long time running with a big number of heap objects.This paper describes a stack-based framework for detecting these “cold” objects at run-time,with a view to marshaling and sequestering them in designated regions of the heap where they may be preferentially paged out to a backing store,thereby freeing physical memory pages for occupation by more active objects.Furthermore,we evaluate the correctness and efficiency of stack-based approach with an Access Barrier.The experimental results from a series of SPECjvm2008 benchmarks are presented.According to the profile result of different standard prove the GC efficiency has 2-5% increase.