• Welcome to Tux Reports: Where Penguins Fly. We hope you find the topics varied, interesting, and worthy of your time. Please become a member and join in the discussions.

Print Spooling vs suspend

J

Jeff Barnett

Flightless Bird
A few days ago we took one of our printers off the network and to a
repair shop. Later we noticed that one of our computers no longer
dropped to sleep (S3 mode). I've spent a few days trying to figure out
why and repair the problem. The computer owner suddenly recalled trying
to print something after the printer was removed from the network so the
print job was in the spool. The print job was deleted and the computer
went to sleep! I'm sending this note for two reasons:

1) In case someone else out there is tracking down a sleep problem -
It's tricky since the printer icon can disappear from the task bar
removing hints about the source of the problem

2) To ask a question based on my memory: MEMORY = (I think that MS
required software to work and play well with sleep modes in order to be
XP certified.) QUESTION = (If memory is correct wouldn't the XP spooler
fail to be XP certified if someone other than MS wrote it?)

Tongue firmly in cheek on 2 above.
--
Jeff Barnett
 
P

Pegasus [MVP]

Flightless Bird
"Jeff Barnett" <jbbrus@ca.rr.com> said this in news item
news:-OoL$IvdpKHA.1548@TK2MSFTNGP06.phx.gbl...
> A few days ago we took one of our printers off the network and to a repair
> shop. Later we noticed that one of our computers no longer dropped to
> sleep (S3 mode). I've spent a few days trying to figure out why and repair
> the problem. The computer owner suddenly recalled trying to print
> something after the printer was removed from the network so the print job
> was in the spool. The print job was deleted and the computer went to
> sleep! I'm sending this note for two reasons:
>
> 1) In case someone else out there is tracking down a sleep problem - It's
> tricky since the printer icon can disappear from the task bar removing
> hints about the source of the problem
>
> 2) To ask a question based on my memory: MEMORY = (I think that MS
> required software to work and play well with sleep modes in order to be XP
> certified.) QUESTION = (If memory is correct wouldn't the XP spooler fail
> to be XP certified if someone other than MS wrote it?)
>
> Tongue firmly in cheek on 2 above.
> --
> Jeff Barnett


I suspect that it is by design that pending print requests should prevent a
PC from going into a sleep state. It is likely that many printers are unable
to cope with an extended interruption in the flow of print data. What would
you say if you had a network printer that refused to accept any more print
jobs because the PC that has just sent half of a print job has been put to
sleep? You would probably slam Microsoft for not having thought this through
properly.

I think you can take your tongue out of your cheek.
 
J

Jeff Barnett

Flightless Bird
Pegasus [MVP] wrote:
>
>
> "Jeff Barnett" <jbbrus@ca.rr.com> said this in news item
> news:-OoL$IvdpKHA.1548@TK2MSFTNGP06.phx.gbl...
>> A few days ago we took one of our printers off the network and to a
>> repair shop. Later we noticed that one of our computers no longer
>> dropped to sleep (S3 mode). I've spent a few days trying to figure
>> out why and repair the problem. The computer owner suddenly recalled
>> trying to print something after the printer was removed from the
>> network so the print job was in the spool. The print job was deleted
>> and the computer went to sleep! I'm sending this note for two reasons:
>>
>> 1) In case someone else out there is tracking down a sleep problem -
>> It's tricky since the printer icon can disappear from the task bar
>> removing hints about the source of the problem
>>
>> 2) To ask a question based on my memory: MEMORY = (I think that MS
>> required software to work and play well with sleep modes in order to
>> be XP certified.) QUESTION = (If memory is correct wouldn't the XP
>> spooler fail to be XP certified if someone other than MS wrote it?)
>>
>> Tongue firmly in cheek on 2 above.
>> --
>> Jeff Barnett

>
> I suspect that it is by design that pending print requests should
> prevent a PC from going into a sleep state. It is likely that many
> printers are unable to cope with an extended interruption in the flow
> of print data. What would you say if you had a network printer that
> refused to accept any more print jobs because the PC that has just
> sent half of a print job has been put to sleep? You would probably
> slam Microsoft for not having thought this through properly.
>
> I think you can take your tongue out of your cheek.

Note that the job had never reached the printer because it was OFF the
network (and at the repair shop). Therefore, the spooler kept trying to
send the job and not even a byte (or a packet depending on how the net
error was reported) of the print stream arrived. The point is that after
N tries, where N was certainly not a small number after two days, there
is reason to believe that 1) the time between tries should have been
backed off and/or 2) a mechanism should have been used that did not
disturb the count-down-to-sleep timer. I'm still of the opinion, given
that my memory is correct, that MS would not have certified this
behavior for another organization. And my tongue is still in my cheek.
--
Jeff Barnett
 
J

Jose

Flightless Bird
On Feb 4, 3:51 pm, Jeff Barnett <jbb...@ca.rr.com> wrote:
> A few days ago we took one of our printers off the network and to a
> repair shop. Later we noticed that one of our computers no longer
> dropped to sleep (S3 mode). I've spent a few days trying to figure out
> why and repair the problem. The computer owner suddenly recalled trying
> to print something after the printer was removed from the network so the
> print job was in the spool. The print job was deleted and the computer
> went to sleep! I'm sending this note for two reasons:
>
> 1) In case someone else out there is tracking down a sleep problem -
> It's tricky since the printer icon can disappear from the task bar
> removing hints about the source of the problem
>
> 2) To ask a question based on my memory: MEMORY = (I think that MS
> required software to work and play well with sleep modes in order to be
> XP certified.) QUESTION = (If memory is correct wouldn't the XP spooler
> fail to be XP certified if someone other than MS wrote it?)
>
> Tongue firmly in cheek on 2 above.
> --
> Jeff Barnett


Do you mean Stand By mode (or Hibernate)?

Why don't you try it again and see.

I unhooked my wireless printer and printed a file. It ended up where
all good print jobs go:

C:/Windows\system32\spool\PRINTERS

As a matter of fact, I printed several things using various programs.

I adjusted my Power Options System standby setting (I could not find
any sleep option) from Never to 2 minutes, OK'd my way out, double
clicked the clock so I could watch the time and in 2 minutes, the
system entered Stand By.

I already know Hibernate works when there are pending jobs in the
spool folder since I choose to use Hibernate instead of Stand By.

I did not get any print icon either so there was nothing to disappear
since it never appeared.

I clicked the mouse to bring the system around, plugged my printer
back in, and was shortly presented a print preview of my document
since that is how I have my system set up (print preview) and the
printer icon showed up until all my print jobs were dispatched. The
files printed fine.

I'm not saying you don't have a problem, but if you can try it and
verify that would be good information to know.

I cannot recreate any situation where a pending print job inhibits
Stand By or Hibernate. If you can think of another way for me to test
it, I will do it.
 
J

Jose

Flightless Bird
On Feb 4, 6:59 pm, Jose <jose_e...@yahoo.com> wrote:
> On Feb 4, 3:51 pm, Jeff Barnett <jbb...@ca.rr.com> wrote:
>
>
>
>
>
> > A few days ago we took one of our printers off the network and to a
> > repair shop. Later we noticed that one of our computers no longer
> > dropped to sleep (S3 mode). I've spent a few days trying to figure out
> > why and repair the problem. The computer owner suddenly recalled trying
> > to print something after the printer was removed from the network so the
> > print job was in the spool. The print job was deleted and the computer
> > went to sleep! I'm sending this note for two reasons:

>
> > 1) In case someone else out there is tracking down a sleep problem -
> > It's tricky since the printer icon can disappear from the task bar
> > removing hints about the source of the problem

>
> > 2) To ask a question based on my memory: MEMORY = (I think that MS
> > required software to work and play well with sleep modes in order to be
> > XP certified.) QUESTION = (If memory is correct wouldn't the XP spooler
> > fail to be XP certified if someone other than MS wrote it?)

>
> > Tongue firmly in cheek on 2 above.
> > --
> > Jeff Barnett

>
> Do you mean Stand By mode (or Hibernate)?
>
> Why don't you try it again and see.
>
> I unhooked my wireless printer and printed a file.  It ended up where
> all good print jobs go:
>
> C:/Windows\system32\spool\PRINTERS
>
> As a matter of fact, I printed several things using various programs.
>
> I adjusted my Power Options System standby setting (I could not find
> any sleep option) from Never to 2 minutes, OK'd my way out, double
> clicked the clock so I could watch the time and in 2 minutes, the
> system entered Stand By.
>
> I already know Hibernate works when there are pending jobs in the
> spool folder since I choose to use Hibernate instead of Stand By.
>
> I did not get any print icon either so there was nothing to disappear
> since it never appeared.
>
> I clicked the mouse to bring the system around, plugged my printer
> back in, and was shortly presented a print preview of my document
> since that is how I have my system set up (print preview) and the
> printer icon showed up until all my print jobs were dispatched.  The
> files printed fine.
>
> I'm not saying you don't have a problem, but if you can try it and
> verify that would be good information to know.
>
> I cannot recreate any situation where a pending print job inhibits
> Stand By or Hibernate.  If you can think of another way for me to test
> it, I will do it.


I also just tried it with a USB wired printer and it works the same
way as my wireless printer - as expected. I wouldn't think MS would
care how the printer was connected and apparently is doesn't in my
situation.
 
J

Jeff Barnett

Flightless Bird
Jose wrote:
> On Feb 4, 6:59 pm, Jose <jose_e...@yahoo.com> wrote:
>
>> On Feb 4, 3:51 pm, Jeff Barnett <jbb...@ca.rr.com> wrote:
>>
>>
>>
>>
>>
>>
>>> A few days ago we took one of our printers off the network and to a
>>> repair shop. Later we noticed that one of our computers no longer
>>> dropped to sleep (S3 mode). I've spent a few days trying to figure out
>>> why and repair the problem. The computer owner suddenly recalled trying
>>> to print something after the printer was removed from the network so the
>>> print job was in the spool. The print job was deleted and the computer
>>> went to sleep! I'm sending this note for two reasons:
>>>
>>> 1) In case someone else out there is tracking down a sleep problem -
>>> It's tricky since the printer icon can disappear from the task bar
>>> removing hints about the source of the problem
>>>
>>> 2) To ask a question based on my memory: MEMORY = (I think that MS
>>> required software to work and play well with sleep modes in order to be
>>> XP certified.) QUESTION = (If memory is correct wouldn't the XP spooler
>>> fail to be XP certified if someone other than MS wrote it?)
>>>
>>> Tongue firmly in cheek on 2 above.
>>> --
>>> Jeff Barnett
>>>

>> Do you mean Stand By mode (or Hibernate)?
>>
>> Why don't you try it again and see.
>>
>> I unhooked my wireless printer and printed a file. It ended up where
>> all good print jobs go:
>>
>> C:/Windows\system32\spool\PRINTERS
>>
>> As a matter of fact, I printed several things using various programs.
>>
>> I adjusted my Power Options System standby setting (I could not find
>> any sleep option) from Never to 2 minutes, OK'd my way out, double
>> clicked the clock so I could watch the time and in 2 minutes, the
>> system entered Stand By.
>>
>> I already know Hibernate works when there are pending jobs in the
>> spool folder since I choose to use Hibernate instead of Stand By.
>>
>> I did not get any print icon either so there was nothing to disappear
>> since it never appeared.
>>
>> I clicked the mouse to bring the system around, plugged my printer
>> back in, and was shortly presented a print preview of my document
>> since that is how I have my system set up (print preview) and the
>> printer icon showed up until all my print jobs were dispatched. The
>> files printed fine.
>>
>> I'm not saying you don't have a problem, but if you can try it and
>> verify that would be good information to know.
>>
>> I cannot recreate any situation where a pending print job inhibits
>> Stand By or Hibernate. If you can think of another way for me to test
>> it, I will do it.
>>

>
> I also just tried it with a USB wired printer and it works the same
> way as my wireless printer - as expected. I wouldn't think MS would
> care how the printer was connected and apparently is doesn't in my
> situation.
>

As I stated in my original message, the sleep/suspend mode was S3;
hibernate is S4. S3 is also called suspend to RAM. As I understand MS
rules for initiating S2/S3/S4, the CPU cannot be more than x% busy for a
user specified period or no state transfer will occur. (Of course you
can give one of the various suspend/hibernate now commands and it will
force transition.) So in part the ability to sleep will depend on the
relative sizes of x%, the user-specified period, the spooler
time-between-checks, the amount of resources necessary to poll the
printer, and several other things. What I managed to discover was that a
printer missing/not responding state can block state transition. The
only time you probably would care about this tidbit of trivia is if your
computer was failing to enter sleep automatically.
--
Jeff Barnett
 
J

Jose

Flightless Bird
On Feb 4, 10:39 pm, Jeff Barnett <jbb...@ca.rr.com> wrote:
> Jose wrote:
> > On Feb 4, 6:59 pm, Jose <jose_e...@yahoo.com> wrote:

>
> >> On Feb 4, 3:51 pm, Jeff Barnett <jbb...@ca.rr.com> wrote:

>
> >>> A few days ago we took one of our printers off the network and to a
> >>> repair shop. Later we noticed that one of our computers no longer
> >>> dropped to sleep (S3 mode). I've spent a few days trying to figure out
> >>> why and repair the problem. The computer owner suddenly recalled trying
> >>> to print something after the printer was removed from the network so the
> >>> print job was in the spool. The print job was deleted and the computer
> >>> went to sleep! I'm sending this note for two reasons:

>
> >>> 1) In case someone else out there is tracking down a sleep problem -
> >>> It's tricky since the printer icon can disappear from the task bar
> >>> removing hints about the source of the problem

>
> >>> 2) To ask a question based on my memory: MEMORY = (I think that MS
> >>> required software to work and play well with sleep modes in order to be
> >>> XP certified.) QUESTION = (If memory is correct wouldn't the XP spooler
> >>> fail to be XP certified if someone other than MS wrote it?)

>
> >>> Tongue firmly in cheek on 2 above.
> >>> --
> >>> Jeff Barnett

>
> >> Do you mean Stand By mode (or Hibernate)?

>
> >> Why don't you try it again and see.

>
> >> I unhooked my wireless printer and printed a file.  It ended up where
> >> all good print jobs go:

>
> >> C:/Windows\system32\spool\PRINTERS

>
> >> As a matter of fact, I printed several things using various programs.

>
> >> I adjusted my Power Options System standby setting (I could not find
> >> any sleep option) from Never to 2 minutes, OK'd my way out, double
> >> clicked the clock so I could watch the time and in 2 minutes, the
> >> system entered Stand By.

>
> >> I already know Hibernate works when there are pending jobs in the
> >> spool folder since I choose to use Hibernate instead of Stand By.

>
> >> I did not get any print icon either so there was nothing to disappear
> >> since it never appeared.

>
> >> I clicked the mouse to bring the system around, plugged my printer
> >> back in, and was shortly presented a print preview of my document
> >> since that is how I have my system set up (print preview) and the
> >> printer icon showed up until all my print jobs were dispatched.  The
> >> files printed fine.

>
> >> I'm not saying you don't have a problem, but if you can try it and
> >> verify that would be good information to know.

>
> >> I cannot recreate any situation where a pending print job inhibits
> >> Stand By or Hibernate.  If you can think of another way for me to test
> >> it, I will do it.

>
> > I also just tried it with a USB wired printer and it works the same
> > way as my wireless printer - as expected.  I wouldn't think MS would
> > care how the printer was connected and apparently is doesn't in my
> > situation.

>
> As I stated in my original message, the sleep/suspend mode was S3;
> hibernate is S4. S3 is also called suspend to RAM. As I understand MS
> rules for initiating S2/S3/S4, the CPU cannot be more than x% busy for a
> user specified period or no state transfer will occur. (Of course you
> can give one of the various suspend/hibernate now commands and it will
> force transition.) So in part the ability to sleep will depend on the
> relative sizes of x%, the user-specified period, the spooler
> time-between-checks, the amount of resources necessary to poll the
> printer, and several other things. What I managed to discover was that a
> printer missing/not responding state can block state transition. The
> only time you probably would care about this tidbit of trivia is if your
> computer was failing to enter sleep automatically.
> --
> Jeff Barnett


Relative sizes, unspecified period - X%, spooler time-between-checks
(what is the spooler time between checks), several other things...
doesn't sound like very good science since none of the variables
mentioned include a method to tie a specific value to one of the
terms. What are the several other things.

Where are you getting these MS rules for initiating S2/S3/S4?

What method was used to discover that a printer missing/not responding
state can block state transition and how would I recreate that state?
I just want to experience it since I hardly ever have my printer on
and have never had a problem with Hibernating that was not self
induced.

I just don't see how a disconnected network or other printer will keep
a system from entering Stand By or Hibernate after the specified
period, assuming it was working before.

If there is a way for merely a disconnected printer to keep my system
from entering Stand By or Hibernate, I would like to know what it is.
I want to somehow tie the inability to enter Stand By to a
disconnected printer.
 
J

Jeff Barnett

Flightless Bird
Jose wrote:
> On Feb 4, 10:39 pm, Jeff Barnett <jbb...@ca.rr.com> wrote:
>
>> Jose wrote:
>>
>>> On Feb 4, 6:59 pm, Jose <jose_e...@yahoo.com> wrote:
>>>
>>>> On Feb 4, 3:51 pm, Jeff Barnett <jbb...@ca.rr.com> wrote:
>>>>
>>>>> A few days ago we took one of our printers off the network and to a
>>>>> repair shop. Later we noticed that one of our computers no longer
>>>>> dropped to sleep (S3 mode). I've spent a few days trying to figure out
>>>>> why and repair the problem. The computer owner suddenly recalled trying
>>>>> to print something after the printer was removed from the network so the
>>>>> print job was in the spool. The print job was deleted and the computer
>>>>> went to sleep! I'm sending this note for two reasons:
>>>>>
>>>>> 1) In case someone else out there is tracking down a sleep problem -
>>>>> It's tricky since the printer icon can disappear from the task bar
>>>>> removing hints about the source of the problem
>>>>>
>>>>> 2) To ask a question based on my memory: MEMORY = (I think that MS
>>>>> required software to work and play well with sleep modes in order to be
>>>>> XP certified.) QUESTION = (If memory is correct wouldn't the XP spooler
>>>>> fail to be XP certified if someone other than MS wrote it?)
>>>>>
>>>>> Tongue firmly in cheek on 2 above.
>>>>> --
>>>>> Jeff Barnett
>>>>>
>>>> Do you mean Stand By mode (or Hibernate)?
>>>>
>>>> Why don't you try it again and see.
>>>>
>>>> I unhooked my wireless printer and printed a file. It ended up where
>>>> all good print jobs go:
>>>>
>>>> C:/Windows\system32\spool\PRINTERS
>>>>
>>>> As a matter of fact, I printed several things using various programs.
>>>>
>>>> I adjusted my Power Options System standby setting (I could not find
>>>> any sleep option) from Never to 2 minutes, OK'd my way out, double
>>>> clicked the clock so I could watch the time and in 2 minutes, the
>>>> system entered Stand By.
>>>>
>>>> I already know Hibernate works when there are pending jobs in the
>>>> spool folder since I choose to use Hibernate instead of Stand By.
>>>>
>>>> I did not get any print icon either so there was nothing to disappear
>>>> since it never appeared.
>>>>
>>>> I clicked the mouse to bring the system around, plugged my printer
>>>> back in, and was shortly presented a print preview of my document
>>>> since that is how I have my system set up (print preview) and the
>>>> printer icon showed up until all my print jobs were dispatched. The
>>>> files printed fine.
>>>>
>>>> I'm not saying you don't have a problem, but if you can try it and
>>>> verify that would be good information to know.
>>>>
>>>> I cannot recreate any situation where a pending print job inhibits
>>>> Stand By or Hibernate. If you can think of another way for me to test
>>>> it, I will do it.
>>>>
>>> I also just tried it with a USB wired printer and it works the same
>>> way as my wireless printer - as expected. I wouldn't think MS would
>>> care how the printer was connected and apparently is doesn't in my
>>> situation.
>>>

>> As I stated in my original message, the sleep/suspend mode was S3;
>> hibernate is S4. S3 is also called suspend to RAM. As I understand MS
>> rules for initiating S2/S3/S4, the CPU cannot be more than x% busy for a
>> user specified period or no state transfer will occur. (Of course you
>> can give one of the various suspend/hibernate now commands and it will
>> force transition.) So in part the ability to sleep will depend on the
>> relative sizes of x%, the user-specified period, the spooler
>> time-between-checks, the amount of resources necessary to poll the
>> printer, and several other things. What I managed to discover was that a
>> printer missing/not responding state can block state transition. The
>> only time you probably would care about this tidbit of trivia is if your
>> computer was failing to enter sleep automatically.
>> --
>> Jeff Barnett
>>

>
> Relative sizes, unspecified period - X%, spooler time-between-checks
> (what is the spooler time between checks), several other things...
> doesn't sound like very good science since none of the variables
> mentioned include a method to tie a specific value to one of the
> terms. What are the several other things.
>
> Where are you getting these MS rules for initiating S2/S3/S4?
>
> What method was used to discover that a printer missing/not responding
> state can block state transition and how would I recreate that state?
> I just want to experience it since I hardly ever have my printer on
> and have never had a problem with Hibernating that was not self
> induced.
>
> I just don't see how a disconnected network or other printer will keep
> a system from entering Stand By or Hibernate after the specified
> period, assuming it was working before.
>
> If there is a way for merely a disconnected printer to keep my system
> from entering Stand By or Hibernate, I would like to know what it is.
> I want to somehow tie the inability to enter Stand By to a
> disconnected printe
>


The spool software periodically, when jobs are pending, checks whether
another job can be initiated. That check uses the CPU, the network if
printers are (supposed to be) on the network, and other port drivers if
printers are (supposed to be) on other ports. The the X in X%, I think,
is around 10. Consider one computer with a 1GHz CPU and another with a
3GHz CPU. Though both will take ROUGHLY the same number of cycles to
execute spooler functions, the faster CPU will take ROUGHLY only one
third the amount of time. The X% is measured as time, not cycles. Of
course there are lots of other periodic and aperiodic tasks performed by
the OS and some of your installed applications. It also seems that some
applications and parts of XP (according to legend) do busy waits on
network and port pokes (poke = data movement or status inquiry). A busy
wait, of course, burns 100% of one CPU core or hyper thread or CPU
depending on architecture. Any of these things as well as others that go
on for more than 10-100 milliseconds can an do reset the
time-to-go-to-sleep counter.

If you want to do an experiment, try the following: In the power
management tab set hard disks off, monitor off, and time to suspend to,
say, 15 minutes. You probably should do this for the administrator
account and your limited user account both. (Note that it is unclear
when, and if, you are at the welcome screen which user's power
management profile is used. I tried to get explicit XP information about
that a year or two ago and at that time none of the MVPs in this
newsgroup seemed to know.) Now take a network printer and take it off
the network. Next, spool a job to that printer.Finally see when and if
your computer goes to sleep. I predict that it will power down the
monitor but will not suspend. Whether you are suspended or not is easy
to tell in S3 because the fans will stop. In S1 they usually continue
spinning making it hard to tell externally what state you are in.

As to the definition of computer states, I suggest that you find a copy
of the ACPI spec, That document, coauthored by Microsoft, Intel,
Toshiba, and one or two other corporations, defines plug and play
standards as well as power management standards. The spec is free for
download and is several hundred pages long. For the most part, it is
readable by software as well as hardware professionals. If you or anyone
else reading this message is deeply interested in how modern computer
components work so well together and what can and does go on in the
hardware abstraction layer of operating systems, this document is
fascinating. Many of you will be genuinely surprised at the amount of
terminology and ideas that are encapsulated in the document that belong
to the world and not a single manufacturer.
--
Jeff Barnett
 
M

MRCS

Flightless Bird
Struggled with this problem for days, and it never would have occurred to me it was a print problem if I hadn't stumbled across it thru System Information.
In my case there was no CPU activity. What may have been the glitch was that I changed the default printer after the incomplete pending print job, not knowing I had one.



Jeff Barnett wrote:

Jose wrote:As I stated in my original message, the sleep/suspend mode was
04-Feb-10

Jose wrote:
As I stated in my original message, the sleep/suspend mode was S3;
hibernate is S4. S3 is also called suspend to RAM. As I understand MS
rules for initiating S2/S3/S4, the CPU cannot be more than x% busy for a
user specified period or no state transfer will occur. (Of course you
can give one of the various suspend/hibernate now commands and it will
force transition.) So in part the ability to sleep will depend on the
relative sizes of x%, the user-specified period, the spooler
time-between-checks, the amount of resources necessary to poll the
printer, and several other things. What I managed to discover was that a
printer missing/not responding state can block state transition. The
only time you probably would care about this tidbit of trivia is if your
computer was failing to enter sleep automatically.
--
Jeff Barnett

Previous Posts In This Thread:

On Thursday, February 04, 2010 3:51 PM
Jeff Barnett wrote:

Print Spooling vs suspend
A few days ago we took one of our printers off the network and to a
repair shop. Later we noticed that one of our computers no longer
dropped to sleep (S3 mode). I have spent a few days trying to figure out
why and repair the problem. The computer owner suddenly recalled trying
to print something after the printer was removed from the network so the
print job was in the spool. The print job was deleted and the computer
went to sleep! I am sending this note for two reasons:

1) In case someone else out there is tracking down a sleep problem -
it is tricky since the printer icon can disappear from the task bar
removing hints about the source of the problem

2) To ask a question based on my memory: MEMORY = (I think that MS
required software to work and play well with sleep modes in order to be
XP certified.) QUESTION = (If memory is correct would not the XP spooler
fail to be XP certified if someone other than MS wrote it?)

Tongue firmly in cheek on 2 above.
--
Jeff Barnett

On Thursday, February 04, 2010 4:10 PM
Pegasus [MVP] wrote:

I suspect that it is by design that pending print requests should prevent aPC
I suspect that it is by design that pending print requests should prevent a
PC from going into a sleep state. It is likely that many printers are unable
to cope with an extended interruption in the flow of print data. What would
you say if you had a network printer that refused to accept any more print
jobs because the PC that has just sent half of a print job has been put to
sleep? You would probably slam Microsoft for not having thought this through
properly.

I think you can take your tongue out of your cheek.

On Thursday, February 04, 2010 4:24 PM
Jeff Barnett wrote:

Pegasus [MVP] wrote:Note that the job had never reached the printer because it
Pegasus [MVP] wrote:
Note that the job had never reached the printer because it was OFF the
network (and at the repair shop). Therefore, the spooler kept trying to
send the job and not even a byte (or a packet depending on how the net
error was reported) of the print stream arrived. The point is that after
N tries, where N was certainly not a small number after two days, there
is reason to believe that 1) the time between tries should have been
backed off and/or 2) a mechanism should have been used that did not
disturb the count-down-to-sleep timer. I am still of the opinion, given
that my memory is correct, that MS would not have certified this
behavior for another organization. And my tongue is still in my cheek.
--
Jeff Barnett

On Thursday, February 04, 2010 6:59 PM
Jose wrote:

Do you mean Stand By mode (or Hibernate)?
Do you mean Stand By mode (or Hibernate)?

Why do not you try it again and see.

I unhooked my wireless printer and printed a file. It ended up where
all good print jobs go:

C:/Windows\system32\spool\PRINTERS

As a matter of fact, I printed several things using various programs.

I adjusted my Power Options System standby setting (I could not find
any sleep option) from Never to 2 minutes, OK'd my way out, double
clicked the clock so I could watch the time and in 2 minutes, the
system entered Stand By.

I already know Hibernate works when there are pending jobs in the
spool folder since I choose to use Hibernate instead of Stand By.

I did not get any print icon either so there was nothing to disappear
since it never appeared.

I clicked the mouse to bring the system around, plugged my printer
back in, and was shortly presented a print preview of my document
since that is how I have my system set up (print preview) and the
printer icon showed up until all my print jobs were dispatched. The
files printed fine.

I am not saying you do not have a problem, but if you can try it and
verify that would be good information to know.

I cannot recreate any situation where a pending print job inhibits
Stand By or Hibernate. If you can think of another way for me to test
it, I will do it.

On Thursday, February 04, 2010 7:02 PM
Jose wrote:

eerI also just tried it with a USB wired printer and it works the sameway as
e
er

I also just tried it with a USB wired printer and it works the same
way as my wireless printer - as expected. I would not think MS would
care how the printer was connected and apparently is does not in my
situation.

On Thursday, February 04, 2010 10:39 PM
Jeff Barnett wrote:

Jose wrote:As I stated in my original message, the sleep/suspend mode was
Jose wrote:
As I stated in my original message, the sleep/suspend mode was S3;
hibernate is S4. S3 is also called suspend to RAM. As I understand MS
rules for initiating S2/S3/S4, the CPU cannot be more than x% busy for a
user specified period or no state transfer will occur. (Of course you
can give one of the various suspend/hibernate now commands and it will
force transition.) So in part the ability to sleep will depend on the
relative sizes of x%, the user-specified period, the spooler
time-between-checks, the amount of resources necessary to poll the
printer, and several other things. What I managed to discover was that a
printer missing/not responding state can block state transition. The
only time you probably would care about this tidbit of trivia is if your
computer was failing to enter sleep automatically.
--
Jeff Barnett

On Friday, February 05, 2010 7:15 AM
Jose wrote:

tngtherbeolerestRelative sizes, unspecified period - X%, spooler
t
ng
the
r
be
oler
e
st

Relative sizes, unspecified period - X%, spooler time-between-checks
(what is the spooler time between checks), several other things...
does not sound like very good science since none of the variables
mentioned include a method to tie a specific value to one of the
terms. What are the several other things.

Where are you getting these MS rules for initiating S2/S3/S4?

What method was used to discover that a printer missing/not responding
state can block state transition and how would I recreate that state?

On Friday, February 05, 2010 1:45 PM
Jeff Barnett wrote:

Jose wrote:
Jose wrote:


Submitted via EggHeadCafe - Software Developer Portal of Choice
Producer/Consumer Queue and BlockingCollection in C# 4.0
http://www.eggheadcafe.com/tutorial...mer-queue-and-blockingcollection-in-c-40.aspx
 
Top