Troubleshooting the parallel PDF RIP client
There are certain factors that can enhance the performance of your pdfprs client. This section provides some recommendations.
- Defining RIPs on an InfoPrint server
- The more RIPs you define, the greater the performance improvement. However, because a single RIP uses an entire CPU, you should never start more RIPs than your InfoPrint Server can support. The pdfprs client generally uses one processor, so be sure that your configuration can support the number of RIPs that you define and start. For guidelines see Processors required to use the Parallel PDF RIP client. For more information about defining RIPs across servers, see RIP Daemon configuration files.
- Specifying an uneven number of RIPs
- If you have PDF files that have simple text data on the front page and more complex image data on the back page consistently throughout the job, a two-RIP configuration will not result in enhanced performance. Since the second RIP always gets the more complex pages, the job will take the same time to RIP as if a single RIP were used, while the first RIP remains idle. Use an uneven number of RIPs to distribute the pages in the job more evenly.
- Using –p or –g options from the command line
- While you can specify a page subset for a particular job by using the –p or the –g flags on the pdfprs command, we recommend that you avoid these flags. If InfoPrint has to determine a subset of the pages in the job, it slows down processing and reduces the benefit of having parallel RIPs.