[cups.general] slow printing from cups

[Ryan Suarez]

Ryan Suarez wrote:
> Michael R Sweet wrote:
>   
>> Ryan Suarez wrote:  
>>     
>>> Greetings,
>>>
>>> We're running samba v3.24 & cups v1.3.7.  The problem queue (oa-e202-e1) 
>>> is printing to the lpd port Lexmark T644 printer.  I'm trying to print 
>>> from Windows Vista 32bit client using  the 'Lexmark T644 PS (MS)' driver 
>>> installed from the clients local driver repository.
>>>
>>> The problem is that it takes a *minute* to print a simple test notepad 
>>> file.  The job hits the printer immediately since the roller on the 
>>> printer starts spinning (slowly).  But the roller keeps spinning 
>>> (slowly) for a while and the job eventually prints (slowly) after a 
>>> minute of waiting.
>>>
>>> If I print directly to the IP of the printer, using the same driver on 
>>> the client, the job prints *right away*.
>>>
>>> I have cups job files and error_log here:
>>> http://it.sheridanc.on.ca/cups
>>>
>>> Any advice?  We're getting alot of complaints with this.
>>>     
>>>       
>> More than likely the issue is that the print file probably gets
>> spooled 4 times - first as a metafile for the printer driver, then
>> from the driver to disk, then over to Samba (which puts a copy on
>> disk), and then finally to CUPS which puts a copy on disk before
>> running the backend which sends it to the printer.
>>
>> You can use IPP (URL = http://cupsserver:631/printers/cupsqueue)
>> to print directly from the Windows system through CUPS to eliminate
>> two of those copies.  
>>     
> Hi Michael,
>
> I'll confirm if bypassing samba and mapping directly to cups IPP speeds 
> up the print.
>
> Unfortunately, thats probably not an option for us.  We take advantage 
> of Samba's point and print tech to support the students here.  They 
> simply map the server on their laptops and double-click on the queue 
> they want to connect to.
>
> Having to train students to manually map each printer they need and 
> choose the correct driver, as they roam to different locations, would 
> prove too cumbersome.
>
> Are there any workarounds I can pursue?
>
> thanks and much appreciated,
> Ryan
>   
yup, printing to cups directly is also fast, arghh...
what do I do now?