Browse Prior Art Database

Original Publication Date: 2000-Aug-01
Included in the Prior Art Database: 2003-Jun-19

Publishing Venue



Disclosed is a method for modeling the performance of a printer management system. Multiple instances of external printer simulations are invoked, versus the single data stream or embedded simulation approaches typical of the prior art. A typical embodiment is depicted in Figure 1. Printer management system 100 receives numerous jobs of varying sizes and complexity. Printer management system 100 transforms a variety of formats into Mixed Object Document Content Architecture (MODCA) [1], which supports sophisticated manipulations. Printer management system 100 converts MODCA into Intelligent Printer Data Stream (IPDS) [2], which provides page level recovery. Printer management system 100 sends IPDS to multiple “virtual” printers 101, 102 and 103. Each of virtual printers 101, 102, and 103 parses but does not rasterize data contained in the IPDS. Virtual printers 101, 102, and 103 may consume IPDS data at different rates. Novel to this embodiment is that each virtual printer is a separate yet lightweight process; with a minimum of test equipment a heterogeneous customer environment of even thousands of printers can be simulated from the point of view of the printer management system, the behavior of which is not altered by the test. This embodiment identifies bottlenecks in production systems, determines prior to production an ideal hardware allocation given a representative sample of input print jobs, and verifies support of new high speed printers by the printer management system before hardware prototypes are developed. Details of the virtual printer are depicted in Figure 2. Attachment layer 200 conducts a dialog with the print management system. One aspect of this dialog is the printer reporting its device attributes to the printer management system. Attachment layer 200 can be configured to report attributes identical to any current or planned model of printer, and may be configured to consume a specified number of pages per minute. Attachment layer 200, rasterizer model 201, and mechanism 202 interact in several ways to model how IPDS data is taken from the printer management system. Attachment layer 200 when run standalone reads IPDS commands up to an End Page command, then sleeps until the fixed time slice for that page has expired. When attachment layer 200 conveys selected IPDS commands for one print page to rasterizer model 201, the model calculates the time required for rasterization and may invoke mechanism model 202 to calculate the time required to fuse toner to the medium. The rasterizer model then sleeps the derived amount of time. Subject to buffers between the printer management system and attachment layer 200, and between attachment layer 200 and models 201 and 202, delays in the virtual printer cause data to back up in the IPDS generator of the printer management system; these blockages model what happens when slower physical printers are attached to the printer management system. Conversely, if a model allocated time slice passes without a sheet of paper completed by mechanism model 202, a data starvation event is identified. On an actual high speed printer data starvation causes destructively sudden deceleration of hardware components. One embodiment for rasterizer model 201 is a multiple regression model of time and space complexity for a page with parameters such as number of bytes on page, number of data objects, rotation and scaling of images, and color saturation. One embodiment for mechanism model 202 is a multiple regression model of time complexity for a sheet side with parameters such as the number of bytes in the compressed rasterized bitmap, the size of the sheet to be printed, and decompression method. 1