![[Intel Navigation Header]](/CONTENT/PIX/HEADER.GIF)
LANSpool Troubleshooting
Contents:
TROUBLESHOOTING BASIC PRINTING PROBLEMS
Troubleshooting Printing Problems
When printing is functioning properly, this is what happens:
The printjob is directed to a parallel port (1, 2, or 3), from either
the application or the command line. (The physical port does not have
to be present in the machine.) The Novell "Capture Command" intercepts
the printjob and routes it out the Network Interface Card to a queue in
the file server. (A queue is nothing more than a special directory for
holding printjobs.)
Once in the queue the printjob goes through three states:
* ADDING
* READY
* ACTIVE
"Adding" indicates that the entire job has not yet been placed in the
queue. The queue will wait for the entire job to be placed in the queue
before the status changes to "Ready."
"Ready" indicates that the entire job is in the queue and ready to be
retrieved and sent to a printer.
"Active" means that the printjob is in the process of being retrieved
and sent to a printer.
The printjobs are retrieved from the queue by the Print Server, which is
prompted to look in the queue by requests from the NODE TSRs. (Each NODE
TSR prompts the Print Server to check a specific queue once every 10
seconds.) When the Print Server finds a printjob in a queue, it sends
the printjob, via the network, to the NODE TSR. The Node TSR simply
directs the printjobs to the appropriate port without changing or
manipulating them.
The printer prints the job!
Problems may occur at any point in the process. Determining where a
printing problem is occurring is half (the FIRST half) of solving the
problem.
To begin, make sure the printjob is getting to the queue intact. The
size and status of the jobs can be checked in PCONSOLE.
If the printjobs do not make it to the queue:
* Make sure the capture command is active and pointing to the correct
queue. (Type in CAPTURE SH at the DOS prompt of the workstation.)
* Make sure your application is sending its jobs to the correct LPT
port (the one that is "captured").
If the printjobs are in the queue but remain in the "adding" state:
Add a timeout parameter to the capture command. (The parameter is
/ti=x, where x=1-1000. A good default value is 10, but higher for
printing graphics.)
If the printjobs are in the queue but remain in the "ready" state:
* Check the "Currently Attached Servers" for that queue. Make sure
the LANSpool print server name appears. If it does not, go to the
"Queue Servers" option for the queue, and insert the LANSpool print
server name.
* Make sure there are no user "holds" on the printjob. (Highlight the
printjob and press ter> to see the detailed status of the job.)
* Check the "Current Queue Status" of the queue. Make sure "Server
can service entries in Queue" is set to "Yes."
If the Currently Attached Server is correct, there are no "Holds" on the
print job, and all settings of the Print Queue Status are set to "Yes,"
go to the print station where the NODE TSR is running (the workstation
where the printer is), and at the DOS prompt (from the LANSPOOL
directory) run STATUS.
If no LANSpool printers are detected:
The NODE TSR needs to be loaded on that workstation. (See the
manual for more information if needed.)
If the status says the printer is "held-off":
1. LANSpool has detected a problem with the port, cable, or printer.
2. Remove the NODE TSR by typing in LSDELETE at the DOS prompt. Then
disable the capture command by typing in ENDCAP and print something
directly to the printer (eg: DIR>LPT1, assuming the printer is on
LPT1). If nothing prints, there is a hardware problem, either with
the port, cable, or printer, that must be resolved before the NODE
TSR will function properly. If it prints fine, reset the printer,
reload the NODE TSR, and printjobs in the queue should start to
print.
If the status says "VAP Connection Failed":
Remove the NODE TSR with the LSDELETE command and then reload it.
If none of the above resolves the problem, it is possible that the NODE
TSR is not requesting printjobs from the correct queue.
Remove the NODE TSR with the LSDELETE command, run NODE NEW and
follow the prompts making sure to select the correct queue. When
prompted to overwrite the existing configuration, respond "Yes."
Printjobs from the queue should start to print.
If these steps fail (for the symptom of printjobs not going from "ready"
to "active" in the queue), call technical support.
If the printjobs go "active" but never leave the queue:
1. Go to the print station and run STATUS. See the above section for
details.
2. If STATUS reports no problems, delete the queue and run BINDFIX.
(See the Novell manuals for details about BINDFIX.) Recreate the
queue, using the same name as before and remembering to assign the
queue to the LANSpool print server.
3. If the problem is widespread in many queues, delete each of the
queues and the print server and run BINDFIX. (See the Novell
manuals for details about BINDFIX.) Then run the LANSpool Setup
utility and recreate the print server, selecting it as the default
print server. Then recreate the print queues, making sure to assign
the queues to the print server.
If the printjobs reach the printer but the printout is very garbled:
1. Place the queue on hold by setting "Print Servers can service
entries in queue" to NO. (This is under "Current Queue Status" in
PCONSOLE/print queue info/queue name/.)
2. Print the same job that previously printed garbled. It will go to
the queue and remain there without printing.
3. Go to "queue information" for that queue in PCONSOLE and get the
Queue ID Number. The queue ID number corresponds to a directory
name under the System directory (ie SYS:SYSTEM\eue id #>).
4. Go to the workstation where the printer is located. LOGIN as
someone that has supervisor rights. Run LSDELETE from the LANSpool
directory. Run ENDCAP from any DOS prompt. Go to the
SYS:SYSTEM\eue id #>. You will see the printjob listed there with
a numeric name. Print that file directly to the printer (ie:
filename>LPT1, assuming the printer is on LPT1).
5. If the printjob prints garbled, the corruption is either caused by
an improper printjob configuration by the application or by the
capture command, or the printer is in the wrong mode (eg,
Postscript instead of PCL).
6. If the printjob prints fine, it is being garbled by the NODE TSR as
it is being passed to the printer.
7. Run NODE NEW. As you follow the prompts, when asked to select
"Polled" or "Interrupt," select the one not previously used. (If
you don't know, try "Interrupt," and if the problem persists, re-do
and try "Polled.")
8. If switching the Polled/Interrupt mode is unsuccessful, try
downloading 350CDRV.EXE (a self-extracting file) from Intel's
online services and using the alternate drivers contained in the
file. (350CDRV.EXE contains new drivers that are compatible with
LANSpool 3.5X, and a README text file with detailed instructions.)
If all of the above is unsuccessful, call technical support to report
the problem.
If random characters of the printjob are being "dropped":
1. At the "Print Station," run LSDELETE from the LANSpool directory.
Next run NODE NEW. When prompted to select between "Polled" and
"Interrupt, choose the option not selected in the original
installation. (If you do not remember, try "Interrupt" first. If
the problem persists, repeat this process and switch to "Polled.")
2. If switching the Polled/Interrupt mode is not successful, try
downloading 350CDRV.EXE (a self-extracting file) from Intel's
online services and using the alternate drivers contained in the
file. (350CDRV.EXE contains new drivers that are compatible with
LANSpool 3.5X and a README text file with detailed instructions.)
If the printjob prints correctly, but VERY slowly:
* If you are using the "Interrupt" setting for the NODE, make sure
that the parallel port supports interrupts, that it is set to the
proper interrupt (int7 for LPT1 and int5 for LPT2), and that there
is not an interrupt conflict with another device (eg, the NIC). If
an interrupt conflict is unavoidable, the NODE must be run in the
"Polled" mode.
* If the printer is on a serial port, make sure the handshaking and
mode settings match the printer and that the port supports the
proper interrupts (int4 for COM1 and int3 for COM2). Check to make
sure there is not an interrupt conflict with another device (eg,
the NIC).
* If there are no interrupt conflicts but printing is still slow,
change the setting from Interrupt to Polled. (At the print station,
run LSDELETE from the LANSpool directory. Then run NODE NEW and
when prompted to select "Polled" or "Interrupt," select "Polled."
* If using the "Polled" setting, increase the speed by setting an
environment variable right before the NODE command in the
autoexec.bat. The command is SET LS_POLLED=XXX. (XXX is a value
between 10 and 600. The default is 150. The higher the value, the
faster the printing--and the slower the workstation while a job is
printing.) You may also want to try changing the setting to
"Interrupt." (To do this, run LSDELETE and then NODE NEW.)
If all of the above does not improve the printing speed to an acceptable
level, call technical support to report the problem.
TROUBLESHOOTING PROBLEMS WITH THE NODE.EXE
Error: Unable to locate the Real interrupt 17 handler.
If this error message appears, it means that another application has
"hooked" interrupt 17. The NODE must have access to this interrupt to
function properly. Make sure that no other TSRs are loaded before the
NODE (except IPX & NETX). NODE will hook interrupt 17 and then release
it so that other applications that need it, can access it.
If the NODE appears to hang when executed:
The NODE has a very long timeout. It is trying to establish connection
with the NLM/VAP. It will not return a failure error message for at
least 60 seconds*
*The length of time could be considerably longer if the SHELL.CFG
has a high SPX ABORT TIMEOUT value.
Eventually the NODE should display a message indicating that the printer
port is assigned as a LANSPOOL printer, OR some kind of error message
should be displayed.
If the workstation is truly "hung", re-run NODE NEW and change the
driver from "Polled" to "Interrupt", or vice-versa.
If it is only "hanging" when the workstation is warm-booted, put a PAUSE
statement in the AUTOEXEC.BAT on the line right before the NODE
statement. The "hanging" is caused by the NODE dropping out of memory
and then being executed (requesting a connection to the NLM/VAP) before
the NLM/VAP recognises that the original connection was broken.
Error: Unable to establish connection with LANSpool....
This occurs when the NODE has sent a packet to the NLM/VAP at the file
server and does not receive a response from it.
* Be sure that the NIC at the file server is not a 3COM 3C501
* Make sure you have the latest Network Shells. As of this writing
NETX 3.26 and IPX 3.10 are the most current. ODI driver 2.0 is
also compatible.
* Remove the SPX ABORT TIMEOUT and IPX RETRY COUNT parameters from
the workstations' SHELL.CFG (unless absolutely needed by another
application).
* Go to the file server and make sure the NLM/VAP is enabled. (use
the LS_STATUS option in the LANSPOOL NLM screen or type in
LS_STATUS at the colon prompt for VAP versions).
* If the NLM/VAP is enabled, select the option LS_TRACK ON (or type
in the command at the colon prompt for VAP users).
* If the "Powerchute" VAP is being used, contact APS to get a later
release that is compatible with LANSpool.
If NODE NEW doesn't detect the LPT ports
The NODE will detect the presence of LPT ports by their DOS
address. When an standard LPT address is active, NODE assumes the
LPT port exists. If you have non-standard LPT cards, ie: cards
that do not use the standard DOS addresses for LPT 1 2 & 3, use the
ALL_LPT switch when executing NODE NEW (and also later, when
running NODE). This will force the NODE to assume that all three
LPT port exist, without relying on the DOS address.
The error: No configuration files for this workstation:
This occurs when the NODE cannot find a configuration file for the
workstation where it is being executed, if the NODE is already loaded on
the workstation, or if the PS=Printserver-name> parameter is incorrect.
1. First run STATUS to see if the NODE is already loaded on the
workstation.
2. If the NODE has not been loaded, run NODE NEW, and when prompted to
save to a configuration file, respond YES. The configuration file
name is the Workstation Address converted into Base 36. It is
created in the LOGIN\LANSPOOL\INI directory.
3. If running NODE NEW and saving to a configuration file still
results in the error message the next time the workstation is
booted and the NODE command run, check the syntax of the NODE
command. It should read:
4. NODE PS=Printserver name> where Printserver-name> is the name of
the default printserver selected in the Lanspool SETUP utility.
5. If you use the ALL_LPT switch when running NODE NEW, you must also
use it when running NODE.
ie:NODE PS=Printserver-name> ALL_LPT
The error: LANSpool has not been loaded on this file server
When the NODE first loads, it listens for a broadcast from the NLM/VAP.
This error message is displayed when it does not hear that broadcast.
1. Use the ps=fileserver-name option with the NETX to make sure the
workstation is attaching to the proper file server.
ie: NETX ps=FS1
2. Check to make sure the NLM/VAP is loaded and active. Whether it is
loaded or not should be immediately apparent by checking the file
server. If it is not loaded, there will not be a LANSPOOL NLM
screen and NetWare 286 file servers will not recognize the command
LS_STATUS when entered at the colon prompt. Use the LS_STATUS
option (or command at the colon prompt) to see if the NLM/VAP is
enabled
3. If the Powerchute VAP is being used, contact APS to get a new
version that is compatible with LANSpool.
4. Make sure the NIC in the file server is not a 3COM 3C501
5. Check the network shells to make sure the latest versions are being
used (at the time of writing NETX 3.26 and IPX 3.10 are the most
current). ODI driver 2.0 is also compatible.
6. Try unloading the LANSpool NLM and reloading it (or downing the
file server and rebooting it to reload the LANSpool VAP).
7. Check FCONSOLE to see if the communication buffers utilization is
equal to the number of buffers available. If so, increase the
number of communication buffers.
TROUBLESHOOTING THE NLM/VAP
Error: Unable to LOGIN as Printserver <Printserver-name>
When the LANSPOOL NLM/VAP loads, it is actually logging into the file
server as a print server. It uses the printserver name and password
much the same as a user uses a user-name and password to login. If the
printserver name and password being used by the NLM/VAP do not match the
printserver name and password in PCONSOLE, NetWare will reject the login
and the error message is returned.
To correct the problem, run the Lanspool SETUP utility. Delete and
Recreate the Printserver (making sure to re-assign the queues to it
once it is re-created) and then select it as the "Default Print
Server". You will see a message stating that the LANSPOOL.NLM or
LANSPOOL.VAP is being copied to the SYS:SYSTEM directory. At this
point you should be able to load the LANSPOOL VAP/NLM successfully.
Note: Contrary to the manual, the NLM and VAP should not be loaded with
the ps=printserver-name parameter.
LANSpool v3.0 and IPX v3.02 when used together may cause the file server
to crash when the NODE is executed. Use IPX v3.10 or later and/or
download the upgrade to version 3.5 of LANSPOOL.
Trademark information