/var/psf/seglist and /var/psf/segments with Input Manager

For InfoPrint Manager actual destinations defined with the PSF DSS, InfoPrint Manager uses /var/psf/seglist and /var/psf/segments (by default) for storing all the AFP, including AFP resulting from the transform of all datatypes (or in InfoPrint Manager terminology, document formats), except ASCII (meaning Proprinter or Quietwriter ASCII, or flat ASCII that cannot be formatted with a page definition through ACIF).

The Input Manager segments incoming jobs once they have been transformed to AFP, or if they come in as AFP, to create smaller pieces. These pieces can then begin to print before a file is entirely transformed, leading to faster first-page-out time. It is more of an advantage for files that need to be transformed (such as PostScript, PCL, PDF, and so forth) than for files that are already in AFPDS or MO:DCA format.

When Input Manager segments the file, it creates a segment list (a list of all the segments) in /var/psf/seglist and the segments themselves, which are by default, 5000K chunks of AFP, in /var/psf/segments. The default segment size of 5000K is probably acceptable for most environments. See PSF (/var/psf/segments) for more information about segment size.

The first segment of a job contains all of the inline resources associated with the job. It is called xxx.seg0. If this segment does not exist for a job, the job does not have any inline resources. This first segment does not follow the segment size configured for the segmenter, that is, it is as large as it needs to be to contain all of the inline resources for the job.

All of the segments have the same name, with the final extension increasing numerically until it reaches the last segment with a name like xxx.segx.end. This segment tells PSF that it does not need to wait for more segments and that the job is finished.

After a job prints and is stacked, its segments are deleted unless the delete-segment-list job attribute is set to false. If segments have been saved or space is needed while processing a job, a cleanup routine deletes the segments that are no longer needed.

Segments are created one after the other and sent to the PSF DSS to improve the processing or printing speed. However, there is a limitation when submitting multiple-document jobs such as multiple AFP files in a single pdpr command or staging to print jobs via hot folders. In this case, the documents are concatenated and placed in the same stream, being seen as a single AFP print file (job).

If the first AFP file contains an inline resource group, then the first segment, xxx.seg0 for the entire multi-document job, will contain this inline resource group and the inline resource groups from the rest of the documents ( the other AFP files) will be discarded. Moreover, if the first document does not contain an inline resource group, the entire concatenated multi-document job will not have an inline resource group xxx.seg0. Therefore, it is strongly recommended to avoid using multi-document jobs having inline resource groups so as not to receive resource error messages.