Operating System - OpenVMS
Showing results for 
Search instead for 
Did you mean: 

Importance of Synchronize-All-Jobs lpd parameter

Respected Contributor

Importance of Synchronize-All-Jobs lpd parameter

Hello all. So... I'm being dense today and I'm reading the explanation of the lpd parameter Synchronize-All-Jobs and not understanding it (I've attached the definition of the parameter for convenience).

My inbound execution queue, TCPIP$LPD_IN_nodename_nn, has been going into the stopped state due to exhaustion of dynamic memory. This happens when pending print jobs pile up due to a stopped destination queue (we're talking about hundreds and hundreds of jobs that can pile up in a matter of minutes.)

I've already upped my quotas to pretty high numbers and I stumbled upon this parameter. It looks like a potential savior to my problem but I don't understand the ramifications of changing it to FALSE, if any. All that's important to me is that my print jobs print in the order they were submitted. I don't care about OPCOM messages about each job completion (as referred to in the parameter description).

Just wondering if anyone with direct experience on this parameter had any input. Thanks!
Honored Contributor

Re: Importance of Synchronize-All-Jobs lpd parameter

LPD is using memory, and not scaling, and the help text implies the folks are aware of the behavior and the limits.

Hundreds and thousands of jobs stuffed into a queue really isn't all that large a value, however.

And yes, setting that to FALSE might paper over the bug. But I'd call up HP directly and discuss this regardless, given this behavior is arguably somewhere between a bug and a design limit within in TCP/IP Services, and you're not really pushing this stuff all that hard.

Alternatively (and if your application permits it), offload this LPD processing onto a box that can handle the load; get VMS out of the path.