Controlling lpd behavior

[Mark Poeschl_2]

We recently upgraded from 5.1A to 5.1B (PK4). I've noticed that the lpd behavior is changed. It now seems to attempt 10 retries of a print job before failing. We manage a couple thousand printers in several remote locations and they get powered off, etc. all the time. Is there some way I can control lpd behavior? I've looked in this forum and the System Admin manual and am not finding anything.

I've attached 'before' and 'after' notice.log entries from a failed print request.

[Mark Sornson]

There is no way to control this aspect of
lpd behavior (other than to bring the printers back on-line).

Is this a request for a feature?

After the syslog message about failing to
connect to the remote host after 11 tries,
does the log file show that the queue daemon
executed "lpc restart", to keep the re-try
behavior going?

As you noticed, prior to 5.B PK4, lpd would
just retry indefinitely. The reason for the
behavior change was that sometimes the
connection was to a print server (or
terminal server) that would get caught in
a 'funny' state, which would not clear with
the prior behavior of continuous retries.

However, when system administrators would
detect the situation and do an "lpc restart"
by hand, the condition would clear. ... So,
the behavior of lpd was changed to imitate
this 'fix by hand', which works in some
cases (as described above).

If the printers are simply powered-off,
there's nothing that lpd can do other than
keep retrying. The "lpc restart" behavior
should appear in the log file.

If you don't see the "lpc restart" behavior,
let us (HP) know.

If you're having a problem connecting to
printers that are powered up, also let us
know.

[Mark Poeschl_2]

I'm not seeing anything in the notice.log file about attempting an 'lpc restart'. What specific message should I be seeing?

[Mark Sornson · Accepted solution]

You should see an INFO log message that says

"openpr - exiting queue daemon with exec of
'lpc restart'"

This message will have the same process id
as the "openpr - failed to connect ..."
message.

After that, a new lpd process should be started
for the queue which retries the job.

What is in the log file after the "failed to
connect ..." message (the last message in
your lpr.log file attachment)?

[Mark Poeschl_2]

You are correct - we had INFO level messages going to the bit bucket. Once I turned them on it works exactly as you've outlined. Thanks for the explanation.

[Mark Poeschl_2]

Behaviour explained...