Printer status "Error - Printing"

We have a couple of Windows Server 2003 (with SP2 applied) that work as a
clustered print server. We have about 35 printers on each server. We use
primarily HP printers (especially 4250 systems).

A couple of times a day, we have a situation where a random set of printers
(about 20 total) will “stop” printing.

Here is what we see:

1. User submits print job
2. It successfully spools to the appropriate printer (we are monitoring the
print spooler pretty much in real time and we open the print queue) in the
spooler.
3. The printer queue status changes to from Spooling to Printing.
4. The job shows in the printer queue as printing for about 2 minutes
(nothing is happening on the printer) and then changes to Error – Printing.
5. The printer in the spooler changes from Ready to Error.
6. The printer display NEVER changes from its “Ready” status (never shows
processing job or other common messages).

The printers are set up:

- to be shared
- the “Enable printer pooling” is not checked
- the “Spool print documents so program finishes printing faster” radio
button is set as well as the Start printing immediately radio button.
- the Print spooled documents first and Enable advanced printing features
boxes are checked

We’ve verified the printer drivers. We’ve verified network connectivity by
pinging the printer IP address (actually the JetDirect card in the printer)
and even opened the printer’s web interface FROM the print server (so we know
that IP connectivity is good).

If the user waits about 20 minutes its “appears” that the Print Server (the
spooler?) tries again and the job goes through. This is the same scenario
regardless if both servers in the cluster are being used or we fail over to
one or the other servers.

The complaint is slow printing.

How can we find out what the “Error” status is when the spooler and the
printer queue show Error?

Is there a way to have the spooler try resubmitting the job on a faster clip.

Why not try setting up a few printers on a stand-alone print server
(non-clustered) to take this out of the equation. Set up a few users on
this new print server printing to the same printers for a few days and see
what happens. I've found clustered print servers to be more trouble than
they are worth in the past....

Cheers,
Jeremy.

"Kdouglas" <Kdouglas@discussions.microsoft.com> wrote in message
news:195066A3-50F9-47BF-ABCB-6837536C7918@microsoft.com...
> We have a couple of Windows Server 2003 (with SP2 applied) that work as a
> clustered print server. We have about 35 printers on each server. We
> use
> primarily HP printers (especially 4250 systems).
>
> A couple of times a day, we have a situation where a random set of
> printers
> (about 20 total) will “stop” printing.
>
> Here is what we see:
>
> 1. User submits print job
> 2. It successfully spools to the appropriate printer (we are monitoring
> the
> print spooler pretty much in real time and we open the print queue) in the
> spooler.
> 3. The printer queue status changes to from Spooling to Printing.
> 4. The job shows in the printer queue as printing for about 2 minutes
> (nothing is happening on the printer) and then changes to Error –
> Printing.
> 5. The printer in the spooler changes from Ready to Error.
> 6. The printer display NEVER changes from its “Ready” status (never shows
> processing job or other common messages).
>
> The printers are set up:
>
> - to be shared
> - the “Enable printer pooling” is not checked
> - the “Spool print documents so program finishes printing faster” radio
> button is set as well as the Start printing immediately radio button.
> - the Print spooled documents first and Enable advanced printing features
> boxes are checked
>
> We’ve verified the printer drivers. We’ve verified network connectivity
> by
> pinging the printer IP address (actually the JetDirect card in the
> printer)
> and even opened the printer’s web interface FROM the print server (so we
> know
> that IP connectivity is good).
>
> If the user waits about 20 minutes its “appears” that the Print Server
> (the
> spooler?) tries again and the job goes through. This is the same scenario
> regardless if both servers in the cluster are being used or we fail over
> to
> one or the other servers.
>
> The complaint is slow printing.
>
> How can we find out what the “Error” status is when the spooler and the
> printer queue show Error?
>
> Is there a way to have the spooler try resubmitting the job on a faster
> clip.
>

I suspect this could be an issue surrounding SP2 to be honest - we've just
updated our SP1 servers (two of which are print servers) to SP2 and all of a
sudden our printing stopped working as it should on both print servers.

- Jobs went into the queue OK and displayed as "printing"
- Nothing else happened until the print spooler was restarted. After that
the queued jobs printed fine (admittedly we didn't leave them 20 minutes or
so as the first poster did).
- Newly submitted jobs after the spooler restart sat there until the spooler
was restarted again.

Now this didn't affect all printers, giving the problem a somewhat random
feel. For instance, our HP units were largely unaffected but the problem
appeared most on our Xerox Phasers.

We tried updating the drivers, re-installing the printers, changing LAN
cables - everything. The devices passed all the usual tests (ping, job sent
directly etc..) but nothing seemed to cure the fault when printing via the
shared printer on the server.

We discovered quite by accident that we could get around it by *enabling*
SNMP in the print device web interface (we don't use it so it is disabled
usually). Jobs then printed fine, on demand as usual, even though SNMP had
been disabled and the device working happily prior to the installation of Svr
2k3 SP2!

The only thing I can attibute it to is the installation of SP2 as everything
worked fine beforehand. It's most bizarre and I can't easily find an
explanation for it, but thought posting it here might help someone else.



Jim

--
I think animal testing is a bad idea - they get all nervous and give the
wrong answers....


"Jeremy" wrote:

> Why not try setting up a few printers on a stand-alone print server
> (non-clustered) to take this out of the equation. Set up a few users on
> this new print server printing to the same printers for a few days and see
> what happens. I've found clustered print servers to be more trouble than
> they are worth in the past....
>
> Cheers,
> Jeremy.
>
> "Kdouglas" <Kdouglas@discussions.microsoft.com> wrote in message
> news:195066A3-50F9-47BF-ABCB-6837536C7918@microsoft.com...
> > We have a couple of Windows Server 2003 (with SP2 applied) that work as a
> > clustered print server. We have about 35 printers on each server. We
> > use
> > primarily HP printers (especially 4250 systems).
> >
> > A couple of times a day, we have a situation where a random set of
> > printers
> > (about 20 total) will “stop” printing.
> >
> > Here is what we see:
> >
> > 1. User submits print job
> > 2. It successfully spools to the appropriate printer (we are monitoring
> > the
> > print spooler pretty much in real time and we open the print queue) in the
> > spooler.
> > 3. The printer queue status changes to from Spooling to Printing.
> > 4. The job shows in the printer queue as printing for about 2 minutes
> > (nothing is happening on the printer) and then changes to Error –
> > Printing.
> > 5. The printer in the spooler changes from Ready to Error.
> > 6. The printer display NEVER changes from its “Ready” status (never shows
> > processing job or other common messages).
> >
> > The printers are set up:
> >
> > - to be shared
> > - the “Enable printer pooling” is not checked
> > - the “Spool print documents so program finishes printing faster” radio
> > button is set as well as the Start printing immediately radio button.
> > - the Print spooled documents first and Enable advanced printing features
> > boxes are checked
> >
> > We’ve verified the printer drivers. We’ve verified network connectivity
> > by
> > pinging the printer IP address (actually the JetDirect card in the
> > printer)
> > and even opened the printer’s web interface FROM the print server (so we
> > know
> > that IP connectivity is good).
> >
> > If the user waits about 20 minutes its “appears” that the Print Server
> > (the
> > spooler?) tries again and the job goes through. This is the same scenario
> > regardless if both servers in the cluster are being used or we fail over
> > to
> > one or the other servers.
> >
> > The complaint is slow printing.
> >
> > How can we find out what the “Error” status is when the spooler and the
> > printer queue show Error?
> >
> > Is there a way to have the spooler try resubmitting the job on a faster
> > clip.
> >
>