[csw-users] cups very slow

Christoph Litauer litauer at uni-koblenz.de
Mon Jan 22 09:53:12 CET 2007

Markus Mayer schrieb:
> On Tuesday, 16 Jan 2007 15:26 +0100, Christoph Litauer wrote:
>> But any action on cupsd [...] is veeery slow. Long timeouts. I tried
>> to truss cupsd, but it doesn't let me ... No messages in error_log
>> ...
>> Any ideas?
> This problem has in fact been tracked down to be caused by
> libgcrypt.so. libgcrypt's 'configure' script assumes that Solaris
> doesn't have /dev/urandom, so it always uses /dev/random.
> CUPS, in turn, uses  libgcrypt.so so set up encrypted connections.
> Therefore, CUPS was always using /dev/random as well.
> So, whenever there wasn't enough entropy in the pool to satify the read
> from /dev/random, the call would block till the entropy pool had been
> re-filled, leading to the 'CUPS is so freaking slow' behavior you and
> others have described.
> Torrey, the maintainer for libgcrypt, was so kind to build a new
> packages within half an hour of me telling him what the problem was.
> These packages are currently available from
> http://www.blastwave.org/testing/libgcrypt-1.2.3,REV=2007.01.21-SunOS5.8-i386-CSW.pkg.gz
> http://www.blastwave.org/testing/libgcrypt-1.2.3,REV=2007.01.21-SunOS5.8-sparc-CSW.pkg.gz

Works perfectly! Thanks a lot for your (and the maintainer of libgcrypt)
very quick help.

... but I still have some problems with the package: My Solaris machine
"learns" the printers of my print server (printhost). So e.g. lpstat -v

device for pdfmail: ipp://printhost:631/printers/pdfmail

If I print a job to pdfmail, the following lpstat works:

ray4:/tmp # env CUPS_SERVER=printhost lpstat -o pdfmail
pdfmail-12491   litauer  27648   Mon Jan 22 09:42:12 2007


ray4:/tmp # lpstat -o pdfmail
ray4:/tmp #

Any ideas?

Christoph Litauer                  litauer at uni-koblenz.de
Uni Koblenz, Computing Center,     http://www.uni-koblenz.de/~litauer
Postfach 201602, 56016 Koblenz     Fon: +49 261 287-1311, Fax: -100 1311
PGP-Fingerprint: F39C E314 2650 650D 8092 9514 3A56 FBD8 79E3 27B2

More information about the users mailing list