I've moved one server box to FC15 to get support for various bits of newer hardware. Aside from preupgrade not working (blowing up near the end and not adding the boot entries) its mostly behaving as a server.
One thing that has totally broken though is the cups printing support.
Its a networked printer (Dell 3115N multi-function) and I've tried it with the 3100 drivers included in Fedora 15 and with the Dell drivers that worked before.
In both cases I see
D [06/Jul/2011:19:01:49 +0100] [Job 2] backendRunLoop(print_fd=0, device_fd=5, s nmp_fd=6, addr=0x7f40de2b8a48, use_bc=1, side_cb=0x7f40ddb48cd0) D [06/Jul/2011:19:01:49 +0100] [Job 2] Read 8192 bytes of print data... D [06/Jul/2011:19:01:49 +0100] [Job 2] Wrote 8192 bytes of print data... D [06/Jul/2011:19:01:49 +0100] [Job 2] Read 8192 bytes of print data... D [06/Jul/2011:19:01:49 +0100] [Job 2] Wrote 8192 bytes of print data... D [06/Jul/2011:19:01:49 +0100] [Job 2] Read 8192 bytes of print data... D [06/Jul/2011:19:01:49 +0100] [Job 2] Set job-printer-state-message to "Unable to write print data: Broken pipe", current level=ERROR D [06/Jul/2011:19:01:49 +0100] [Job 2] Print file sent, waiting for printer to f inish... D [06/Jul/2011:19:01:49 +0100] [Job 2] Backend returned status 1 (failed) D [06/Jul/2011:19:01:49 +0100] [Job 2] Printer stopped due to backend errors; pl ease consult the error_log file for details. D [06/Jul/2011:19:01:49 +0100] [Job 2] End of messages
But no real further info anywhere else which is making it a bit hard to nail down what is going on.
I can print happily from an FC14 box on the same network.
Hi Alan,
On Fri, 2011-07-08 at 13:06 +0100, Alan Cox wrote:
D [06/Jul/2011:19:01:49 +0100] [Job 2] backendRunLoop(print_fd=0, device_fd=5, snmp_fd=6, addr=0x7f40de2b8a48, use_bc=1, side_cb=0x7f40ddb48cd0) D [06/Jul/2011:19:01:49 +0100] [Job 2] Read 8192 bytes of print data... D [06/Jul/2011:19:01:49 +0100] [Job 2] Wrote 8192 bytes of print data... D [06/Jul/2011:19:01:49 +0100] [Job 2] Read 8192 bytes of print data... D [06/Jul/2011:19:01:49 +0100] [Job 2] Wrote 8192 bytes of print data... D [06/Jul/2011:19:01:49 +0100] [Job 2] Read 8192 bytes of print data... D [06/Jul/2011:19:01:49 +0100] [Job 2] Set job-printer-state-message to "Unable to write print data: Broken pipe", current level=ERROR D [06/Jul/2011:19:01:49 +0100] [Job 2] Print file sent, waiting for printer to finish... D [06/Jul/2011:19:01:49 +0100] [Job 2] Backend returned status 1 (failed) D [06/Jul/2011:19:01:49 +0100] [Job 2] Printer stopped due to backend errors; please consult the error_log file for details. D [06/Jul/2011:19:01:49 +0100] [Job 2] End of messages
The socket backend is getting EPIPE when trying to send data to the printer. Probably the printer sent a RST?
Here's a wiki page about debugging printing problems: https://fedoraproject.org/wiki/How_to_debug_printing_problems
Tim. */
The socket backend is getting EPIPE when trying to send data to the printer. Probably the printer sent a RST?
Here's a wiki page about debugging printing problems: https://fedoraproject.org/wiki/How_to_debug_printing_problems
I finally pinned this down with some leads from that. It was not what I had expected at all.
The 3115CN was seeing some packet loss from a dodgy network port and it seems that while everything else on the network was quite happy with the losses it was enough to freak the 3115CN's TCP stack. With that sorted it's running out of the box with FC15.
Alan PS: Telsa says hi...
On Wed, 2011-08-03 at 19:42 +0100, Alan Cox wrote:
The 3115CN was seeing some packet loss from a dodgy network port and it seems that while everything else on the network was quite happy with the losses it was enough to freak the 3115CN's TCP stack.
!! A weird one.
Alan PS: Telsa says hi...
:-) Hi back...
Tim. */
-
Alan Cox -
Tim Waugh