strange message in lpr.log

[Michael Schulte zur Sur]

Hi,

I am using extended printserver and access them from Tru Unix 5.1A by TCP/IP printing. I get the following messages in lpr.log.
Sep 17 18:59:31 gre10d lpd[52868]: WARNING -- tcp_conn: trying to connect to gre
056065, error: Connection refused

does anyone have an idea, where to look for?

Michael

[Joris Denayer]

Michael,

Can you give the printer entry in /etc/printcap ?
Can you post also an exact snippet of the messages in lpr.log
What is the output of the "lpc stat" command and of "lpq -P " ?

Is the printer reachable (pingable) over the network ?

Sometimes the lpd section of sys_check gives also useful information.

Also interesting info will be available if you change in /etc/syslog.conf: lpr.info --> lpr.debug.
And afterwards a
# kill -HUP

Hope this helps

Rgrds

Joris

[Ralf Puchner]

This "Connection refused" warning isn't necessarily a real problem. It just takes a
few seconds to reset between the time a previous connection is closed and a new one is allowed to open. Lpd just keeps trying until the connection can be made.

Btw. is the printjob printed or stalled?
Please post the whole log file of the printer if no paper will be printed.

[Michael Schulte zur Sur]

Hi Joris,

first the printcap entry+lpc status
second, the output in lpr.log.

gre056067m1: :lf=/usr/spool/printerlogs/gre056067m1: :lp=@gre056067/lpt1: :of=/usr/lbin/myof: :sd=/usr/spool/gre056067m1: :sh:

printer is on device '@gre056063/lpt1' speed -1
queuing is enabled
printing is enabled
no entries
no daemon present


Sep 18 13:38:44 gre10d lpd[100688]: DEBUG -- openpr: opening TCP connection, printer gre056063m1
Sep 18 13:38:44 gre10d lpd[100688]: DEBUG -- Entering tcp_open()
Sep 18 13:38:44 gre10d lpd[100688]: DEBUG -- Entering tcp_conn()
Sep 18 13:38:44 gre10d lpd[100688]: WARNING -- tcp_conn: trying to connect to gre056063, error: Connection refused
Sep 18 13:38:45 gre10d lpd[100688]: DEBUG -- writeStatus: status file update, printer gre056063m1: Thu Sep 18 13:38:45 2003: gre056063m1 is ready and printing

Hi Ralf,

I don??t know about the printing. It is at another location. I assume, if it didn??t work, I would already have them complain. I haven??t seen this type of message from 4.0F, so I am wondering. It has all started with 5.1A.
Could it be, that the printserver is just busy?

thanks for taking the time,

Michael

[Joris Denayer]

Michael,

As far as I can see, everything is printed correctly.

There are indeed some algorithms changed in lpd. And they have mostly to do with the way timeouts are handled.
Apparently, this warning occurs mostly in combination with printers connected to terminal severs and printservers.

As long as this line occurs once for a givven printjob, there is no need to panic.

Rgrds

Joris

[Ralf Puchner]

Please try JJ#0 as a parameter within printcap. In 5.x the behaviour how a connection will be made changed.
(read the man page of printcap for a description)

With the above parameter the behaviour changes to the old (4.x) one.

[Joris Denayer]

Ralf,

It might help, but according the descriptions, the jj flag is needed when you have multiple jobs to print on the same printer. The timeout between the different jobs will be shortened.
fi: If you have an application that print hundreds of small jobs, where each jobs takes 1 or 2 seconds to print, then the extra 5 second timeout between the jobs will be shortened.
Also, this flag was already entered in v4.0f (introduced in patchkit#6)

Rgrds

Joris

[Ralf Puchner]

Joris,

the description contains no detailed description and no information what is really going on.

The extraction from the log contains only a warning - an explanation for the message was given in my first posting. So my next question was if the printjob will not be printed. If not it is a communication problem if yes we can safely ignore the warning.

The last "guess" was it depends on some changes and this features was one major change in the code (except the cluster stuff). The features was also introduced in 4.0f patchkit #6 but was backported after version 5.0 was instroduced. From the given information it is not clear if the warning depend on one or more printjobs a time...

regards

Ralf

[Michael Schulte zur Sur]

Hi everybody,

actually there can be more than one lpd with different pids found. To avoid confusion, I selected one thread. I just reassured myself, that printing works fine and so I assume, that it is not a real problem but more like collisions on a network.

Thanks for everyone participating.

Michael