Collecting performance metrics from java virtual machines

1. A method of collecting performance metrics data from Java virtual machines, comprising:

setting up a virtual storage structure of a collector on a computing system for collecting performance metrics data from a plurality of Java virtual machines;

pushing, at each of the plurality of Java virtual machines through a corresponding performance monitoring Java agent, performance metrics data collected by the Java virtual machine to the virtual storage structure of the collector;

pulling, at a performance monitoring system through a collector application programming interface (API), the performance metrics data collected by the plurality of Java virtual machines from the virtual storage structure of the collector;

analyzing the performance metrics data pulled from the virtual storage structure of the collector by the performance monitoring system; and

generating, by the performance monitoring system, a performance alert when the performance metrics data analyzed indicates one or more system abnormalities; wherein setting up a virtual storage structure comprises:

setting up a global vector table (GVT) and structures for storing performance metrics data collected from each of the plurality of Java virtual machines, a program call (PC) routine, and a system-wide addressable name-token with a unique name comprising a collector identifier (ID), and a token having a PC number, a GVT address and an address space token (STOKEN).

2. The method of claim 1 , wherein pushing performance metrics data comprises:

pushing the performance metrics data collected by the Java virtual machine to the virtual storage structure of the collector through a low-level cross-memory service.

3. The method of claim 1 , wherein pushing performance metrics data comprises:

pushing the performance metrics data collected by the Java virtual machine to the virtual storage structure of the collector continuously.

4. The method of claim 1 , wherein pulling performance metrics data comprises:

pulling the performance metrics data collected by the plurality of Java virtual machines from the virtual storage structure of the collector periodically in a predetermined interval.

5. The method of claim 1 , wherein pulling performance metrics data comprises:

a historical collection;

a situation collection; and

an interactive collection.

6. The method of claim 1 , wherein the performance metrics data comprises:

garbage collection performance;

threads and locking details;

system response time;

system throughput;

central processing unit (CPU) utilization;

memory utilization; and

Java virtual machine environment details.

7. A computing system comprising a memory storing computer-executable instructions for the computing system, and a processor for executing the computer-executable instructions, the computer-executable instructions comprising:

setting up a virtual storage structure of a collector on the computing system for collecting performance metrics data from a plurality of Java virtual machines;

pushing, at each of the plurality of Java virtual machines through a corresponding performance monitoring Java agent, performance metrics data collected by the Java virtual machine to the virtual storage structure of the collector;

pulling, at a performance monitoring system through a collector API, the performance metrics data collected by the plurality of Java virtual machines from the virtual storage structure of the collector;

analyzing the performance metrics data pulled from the virtual storage structure of the collector by the performance monitoring system; and

generating, by the performance monitoring system, a performance alert when the performance metrics data analyzed indicates one or more system abnormalities; wherein setting up a virtual storage structure comprises:

setting up a global vector table (GVT) and structures for storing performance metrics data collected from each of the plurality of Java virtual machines, a program call (PC) routine, and a system-wide addressable name-token with a unique name comprising a collector identifier (ID), and a token having a PC number, a GVT address and an address space token (STOKEN).

8. The computing system of claim 7 , wherein pushing performance metrics data comprises:

pushing the performance metrics data collected by the Java virtual machine to the virtual storage structure of the collector through a low-level cross-memory service.

9. The computing system of claim 7 , wherein pushing performance metrics data comprises:

pushing the performance metrics data collected by the Java virtual machine to the virtual storage structure of the collector continuously.

10. The computing system of claim 7 , wherein pulling performance metrics data comprises:

pulling the performance metrics data collected by the plurality of Java virtual machines from the virtual storage structure of the collector periodically in a predetermined interval.

11. The computing system of claim 7 , wherein pulling performance metrics data comprises:

a historical collection;

a situation collection; and

an interactive collection.

12. The computing system of claim 7 , wherein the performance metrics data comprises:

garbage collection performance;

threads and locking details;

system response time;

system throughput;

CPU utilization;

memory utilization; and

Java virtual machine environment details.

13. A computer program product for collecting performance metrics data from Java virtual machines, comprising a computer-readable storage medium having computer-executable instructions embodied therewith, when executed by a processor on the computing system, the computer-executable instructions cause the processor to perform:

setting up a virtual storage structure of a collector on the computing system for collecting performance metrics data from a plurality of Java virtual machines;

pushing, at each of the plurality of Java virtual machines through a corresponding performance monitoring Java agent, performance metrics data collected by the Java virtual machine to the virtual storage structure of the collector;

pulling, at a performance monitoring system through a collector API, the performance metrics data collected by the plurality of Java virtual machines from the virtual storage structure of the collector;

analyzing the performance metrics data pulled from the virtual storage structure of the collector by the performance monitoring system; and

generating, by the performance monitoring system, a performance alert when the performance metrics data analyzed indicates one or more system abnormalities; wherein setting up a virtual storage structure comprises:

setting up a global vector table (GVT) and structures for storing performance metrics data collected from each of the plurality of Java virtual machines, a program call (PC) routine, and a system-wide addressable name-token with a unique name comprising a collector identifier (ID), and a token having a PC number, a GVT address and an address space token (STOKEN).

14. The computer program product of claim 13 , wherein pushing performance metrics data comprises:

pushing the performance metrics data collected by the Java virtual machine to the virtual storage structure of the collector through a low-level cross-memory service; and

pushing the performance metrics data collected by the Java virtual machine to the virtual storage structure of the collector continuously.

15. The computer program product of claim 13 , wherein pulling performance metrics data comprises:

pulling the performance metrics data collected by the plurality of Java virtual machines from the virtual storage structure of the collector periodically in a predetermined interval.

16. The computer program product of claim 13 , wherein pulling performance metrics data comprises:

a historical collection;

a situation collection; and

an interactive collection.

17. The computer program product of claim 13 , wherein the performance metrics data comprises:

garbage collection performance;

threads and locking details;

system response time;

system throughput;

CPU utilization;

memory utilization; and

Java virtual machine environment details.