Browse Prior Art Database

Persistent Process Model

IP.com Disclosure Number: IPCOM000176066D
Original Publication Date: 2008-Nov-04
Included in the Prior Art Database: 2008-Nov-04
Document File: 1 page(s) / 19K

Publishing Venue

IBM

Abstract

System to generate and maintain a collection of general purpose UNIX processes.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 100% of the total text.

Page 1 of 1

Persistent Process Model

Disclosed is a method to create and maintain a UNIX process collection for reuse by a program. When a UNIX program creates a process using the fork system call the OS replicates all mutexes, condition variables, virtual memory, file handles and other kernel resources into the new process. When a process is purposed for general use and is not required to inherit the callers kernel resources then the process should be created during a period when the parent process has little activity. This ensures each child process inherits minimal kernel resource and thereby reduces the resource required to maintain the newly forked child process.

After accumulating a finite collection of general purpose processes, the parent process can select an available general purpose child process to start a general purpose program for execution on the selected child process. When the child process ends it's program and avoids exiting to the OS, the parent process can reuse the child process for subsequent general purpose work.

This approach of creating and maintaining persistent processes avoids the situation where a heavily loaded parent process blocks execution on a fork system call when creating a child process. By maintaining a queue of general purpose processes that are created during little or no activity periods of the parent process, the resulting child processes inherit minimal kernel resource and the parent can re...