tom@wcc.oz (Tom Evans) (04/21/88)
Re; Paul Clayton - The TSO Financial Saga, Chapter 34 - April 9, 1988 I'm trying to remember the exact details of the relevant comments as I type this on an RT-11 system to be shipped to you via Kermit and a friendly UN*X system. I was talking to an acquaintance at a DECUS leadership meeting about the problem of clusters, LAT and printer queues becoming tangled. He had been to a USA DECUS sysmposium (Lucky man!) and had talked to a designer/programmer of LAT software. The LAT guru said (not Quoting): - Run one queue on one cpu for each printer. A problem with one generic queue driving "N" execution queues onto one server port is that the execution queues take it in turn to drive the server port and it takes a few seconds to connect/disconnect via LAT. - JBCSYSQUE.DAT was fragile and did not like being hammered. I can't remember any further comments. The less queues the better. Doing a START/QUEUE/MANAGER/NEW can help greatly. - There is a bug in the STOP/QUEUE/RESET command. One way to get rid of an aborting print job is use the above command. Sometimes when the print queue is restarted, the message SYSTEM-F-ILLIOFUNC(?) appears. Look it up for more details. I'm sorry I can't remember more details. The Company I work for has about 100 printers on Ethernet being driven by 2 x VAX 8700 running VMS 4.6. That cluster has similar queue problems to yours. My VAX 8530 (VMS 4.6 about to go to 4.7) only has about 30 of these defined and gets into far less trouble. Andrew Stewart Mad Hobbyist and DECaholic.