Re: 4GL and writing reports to a report

Ed Schaefer wrote: > > Hi Informix professionals: > > Under SCO, I'm writing to a pipe such as: > > START REPORT TO PIPE lp_command > > where lp_command equals something like > > rcmd <node_name> lp <comand flags> <printer_name> > > So far, nothing out of the oridinary. My customer reports that > performance suffers when a certain report is executed. The fix is to > cycle the power on the printer pointed to by <printer_name> and > performance returns to normal. > > Now, I've heard of the unix print que getting so messed up that > rebooting the operating system is necessary, but never just cycling a > printer. > > I was wondering if anyone else has ever seen such a phenomena. Any > insight would be appreciated. > > regards, > > Ed > Schaefer Some random thoughts culled from years of fighting printers... How is the printer connected? If via a parallel port, do you use polling or interrupts to service printer requests? If polling (required on some versions of Xenix on some hardware), that could be the problem. Is the printer sending messages back to the port? If it is a serial port, are you using xon/xoff control? If so, try out-of-band signalling with rts/cts or some such. Is the print job asking the printer to do some complex processing which is bogging the system down with printer wait signals? This would apply to a PostScript printer. Is it an ancient dot-matrix? Could you change the buffer size or high-water mark to fit better with the data speed? Could you use the sar program to get more information? -- Peter Lancashire Information Systems Specialist, Bayer plc Eastern Way, Bury St Edmunds, Suffolk, IP32 7AH, UK Tel: +44-1635-562258, Fax: +44-1635-562281 --- If all else fails, read the instructions. All opinions are my own and not those of Bayer plc. My Internet plumbing does not allow me to mail and post news together. Sorry. --- Join Infuse, the UK Informix User Group at http://www.infuse.org.uk/