Thanks all guys that has replied to me.
---------- Forwarded Message ----------
From: Alan Sherlock, INTERNET:ajs_at_cxo.dec.com
TO: (unknown), [71055,111]
DATE: 6/4/97 11:10 AM
RE: RE: SUMMARY: lpr problem
Sender: ajs_at_cxo.dec.com
Received: from mail12.digital.com (mail12.digital.com [192.208.46.20]) by dub-img-4.compuserve.com (8.6.10/5.950515)
id KAA16071; Wed, 4 Jun 1997 10:48:45 -0400
Received: from tsc.csc.cxo.dec.com by mail12.digital.com (8.7.5/UNX 1.5/1.0/WV)
id KAA18506; Wed, 4 Jun 1997 10:39:33 -0400 (EDT)
Received: from talon.osn.cxo.dec.com by tsc.csc.cxo.dec.com (5.65v4.0/1.1.10.5/16May97-1038PM)
id AA03602; Wed, 4 Jun 1997 08:39:28 -0600
Received: by talon.osn.cxo.dec.com with Microsoft Mail
id <01BC70C3.D139EFE0_at_talon.osn.cxo.dec.com>; Wed, 4 Jun 1997 08:46:39 -0600
Message-Id: <01BC70C3.D139EFE0_at_talon.osn.cxo.dec.com>
From: Alan Sherlock <ajs_at_cxo.dec.com>
To: "'71055.111_at_CompuServe.COM'" <71055.111_at_CompuServe.COM>
Subject: RE: SUMMARY: lpr problem
Date: Wed, 4 Jun 1997 08:46:36 -0600
Encoding: 165 TEXT
Final summary for my lpr problem, thanks for all the answers.
This is a known problem on v3.2c. There is a patch to lpd that fixes it. Contact the CSC. The patch id is OSF350-172.
Alan Sherlock
Digital Customer Support Center
UNIX Networking Support
1-800-354-9000
71055.111_at_CompuServe.COM wrote:
>
> Hi,
>
> Thanks for your answers specially to Dejan Muhamedagic.
>
> I got a partial summary of my lpr problem.
>
> I receive the following from Dejan Muhamedagic:
>
> Hello,
>
> This is silly problem, because it seems that there is a problem with
> filename generation. Apparently, DEC's lpr generates the job number and the
> spool filename accordingly like this:
>
> ----------------------------------------------------------------
> #lpr -Plp4 -j /etc/hosts
> Job number is:51
> #ls
> .seq dfA051sal.yunix.bits.net status
> cfA051sal.yunix.bits.net lock
>
> -------------------------------------------------------------
>
> Perhaps you can try with freeware distribution of lpr
>
> My original question:
>
> Dear Friends:
>
> We have an Alpha Server 2100 with two 250 processors, 256 Mbytes RAM and 8
> GBytes in RAID 5
> Array. We are running DU V3.2C ( Rev 148 ).
>
> We run an application that generates a lot of statements daily that are
> sent by lpr to print servers in our entire net. The problem arise at the
> end of the month when around 4500-5000 statements are generated and we
> realized that some of the statements are lost. We follow the mechanism and
> ensure that the application passes the correct number of statements to unix
>
> for printing.
>
> The only way that we don't lose any job is if I stop the application before
> the job number reachs 1000 ( I see that with the command "lpc status
> <printer_name >"), wait for the queue to be empty and start the printing
> in this remote printer.
>
> I do this every month. :-(
>
> The problem is when the job number is over 1000. I made a shell script
> that send the a file to remote printer because I want to see was was wrong.
>
> ########################################################################
>
> #!/sbin/sh
> COUNT=0
>
> while [ $COUNT != 2000 ]
> do {
> echo "Job number is $COUNT"
> COUNT=`expr $COUNT + 1`
> lpr -h -P remoto test_file
> };
> done
>
> ########################################################################
>
> When the job number is over 1000 the system displays the following error
>
> ########################################################################
>
> Job Number is 997
> Job Number is 998
> Job Number is 999
> Job Number is 1000
> lpr: cannot create /usr/spool/lpd20/dfA183bfibackup
> Job Number is 1001
> lpr: cannot create /usr/spool/lpd20/dfA184bfibackup
> Job Number is 1002
> lpr: cannot create /usr/spool/lpd20/dfA185bfibackup
> Job Number is 1003
> lpr: cannot create /usr/spool/lpd20/dfA186bfibackup
> Job Number is 1004
> lpr: cannot create /usr/spool/lpd20/dfA187bfibackup
> Job Number is 1005
> lpr: cannot create /usr/spool/lpd20/dfA188bfibackup
> Job Number is 1006
> lpr: cannot create /usr/spool/lpd20/dfA189bfibackup
> Job Number is 1007
> lpr: cannot create /usr/spool/lpd20/dfA190bfibackup
> Job Number is 1008
> lpr: cannot create /usr/spool/lpd20/dfA191bfibackup
>
> And automatically the queue becomes to decrease, the jobs is lost too.
>
> In few words
>
> 1. The printer lose the statement intermittently,
> 2. The majority of the statement file size is roughly 6 Kbytes.
> 3. The disk space is enough. The spool directory is located in the "/usr"
> filesystem.
> This filesystem is configured with 2 Gbytes for data storage. This is the
> output
> of df command.
>
> df -k
>
> Filesystem 1024-blocks Used Avail Capacity Mounted on
> /dev/re0a 96079 52825 33646 61% /
> /proc 0 0 0 100% /proc
> /dev/re0g 2034671 1178232 652971 64% /usr
> /dev/re0h 4826573 3377501 966414 78% /u1
>
> If one statement is aprox. 6 Kbytes, in the worst case, we will print 5000
> statements and the disk space occupied by all jobs to be printed will be ~
> 30 Mbytes.
>
> 4. I use a print server configured with the TCP/IP suite of protocols.
> A copy of "/etc/printcap" is supplied.
>
> axis|main_print_server-pr| Network printer on Print server
> main_print_server:\
> :lp=:\
> :sd=/var/spool/lpd/main_print_server-pr:\
> :lf=/var/spool/lpd/main_print_server-pr/log:\
> :rm=main_print_server:\
> :rp=pr1:
>
> 5. The situation happened when print jobs exceeded 1000 in the queue.
>
> 6. We use the Digital command "lprsetup" because is very ease the process
> for setup a printer. In all ours cases, all the printer are remote and
> two parameters needs lprsetup for configuring this type of printer.
>
> Rm : Remote system ( found in /etc/hosts )
> Rp : Remote printer ( the name of the queue in the print server )
>
> We use the default values for the rest of the options.
>
> I think the maximum number of jobs lpr can control up to 1000 but how can
> we
> overcome this limitation. Today, in the world, there are several unix
> applications that send more than 1000 jobs to be printed.
>
> I revised the lpr man page ( and all entries in /etc/printcap ) and there
> are not any parameter in relation with the job's number.
>
> Do you know if there is any kernel configuration parameter for controlling
> the number of the lpr queue?
>
> Thanks in advance for the cooperation I'm sure you will give to this matter
>
> Abdon Tremols
>
>
Received on Wed Jun 11 1997 - 19:24:08 NZST