[cups.general] Maximum number of printer queue's

Johannes Meixner jsmeix at suse.de
Fri Apr 18 03:09:08 PDT 2008


Hello,

On Apr 18 11:28 Johannes Meixner wrote (shortened):
> On Apr 18 00:46 Kjell wrote (shortened):
> > Also I am happy to hear that cups is usable for large installation,
> > since the Novell pages claims that cups is only usable for small
> > to medium sized print environment, and hence you should go
> > for iPrint that can support several hundreds(?) of printer
> > queues :-)
> 
> Could you provide me an URL where you got this information?

I found this document
http://www.novell.com/collateral/4621412/4621412.pdf
which reads
------------------------------------------------------------------
While CUPS will meet many of the needs of small to medium
printing environments, it does not scale to meet the needs
of enterprise customers. iPrint, on the other hand, can host
hundreds of printers and process substantial amounts of data
that CUPS cannot approach.
With CUPS, the conversion into the printer-specific format
of all data to be printed takes place by default directly
on the server.
The advantage is that a print client system that does
filtering is not necessary; the disadvantage is that
the filtering process devours too many resources when
a single server must support hundreds of printers.
With iPrint, in contrast, the filtering takes place
on the client system, but this requires the installation
of printer drivers on the client.
If you are setting up printing in an enterprise environment,
check on iPrint availability before making a final decision. 
------------------------------------------------------------------

Is this what you read?

If the unconditioned "that CUPS cannot approach" wording
was replaced by "that CUPS doesn't approach by default",
I would not see a problem because I think the pro and cons
are sufficiently described (in such a short text).

Of course they don't say explicitely that with CUPS
you can use raw queues on the server and do the filtering
on the clients if you like while with iPrint you cannot
do the filtering on the server so that with iPrint you must
install printer driver software on each client system
and maintain it for each client system.

Perhaps even with raw queues a CUPS server might be slower
than an iPrint server - perhaps noticeable only above a
certain number of print queues or simultaneously active
print jobs - perhaps only with CUPS 1.1 or whatever?
Currently I don't have comparison data for this case.


Kind Regards
Johannes Meixner
-- 
SUSE LINUX Products GmbH, Maxfeldstrasse 5, 90409 Nuernberg, Germany
AG Nuernberg, HRB 16746, GF: Markus Rex





More information about the cups mailing list