102408-a-way-to-hold-completed-jobs?

[ed smith]

Greetings,

Is there an already established way to place large numbers of **completed** jobs on HOLD for a specific cups printer quickly, in a crisis situation? The browser only presents a RESTART button for each completed job, until a new job needs the slot (we run maxjobs at 2500) and bumps the completed job into the bit bucket.

As an example of where this would be useful- Occasionally our batch cycle will print 600 reports to 5000 unreadable pages because of maddening flukes like having little or no ink, usually when no human printer operator is present to stop the bleeding.

We often have to resort to going back to running a many hour report generation batch cycle, or cherry picking our way thru our Oracle report queue, or even the CUPS Q, which is time consuming and does not meet deadlines. It is imperative that the recovery technique produces the reports in the specific original order, so that makes using the original CUPS queue desirable.

I have also found that using command line responses are time consuming to hack together, ie, lpstat -W completed -o | grep -i <my-printer>, and then derive file name lists to put into scripts to manipulate the "c" or "d" files in /var/spool/cups. A single command that can HOLD a subset of completed jobs would be golden.

any suggestions appreciated. Thanks
Ed Smith