On Thu, 2008-05-08 at 14:58 -0500, Gregory P. Ennis wrote:
> On Thu, 2008-05-08 at 12:31 -0400, Ross S. W. Walker wrote:
> > Gregory P. Ennis wrote:
> > > On Thu, 2008-05-08 at 10:46 -0400, Ross S. W. Walker wrote:
> > > > Gregory P. Ennis wrote:
> > > > >
> > > > > On Wed, 2008-05-07 at 12:00 -0500, Gregory P. Ennis wrote:
> > > > > > Everyone,
> > > > > >
> > > > > > I have a new installation of Centos 5.1 that I am using on a gateway
> > > > > > server that also has dhcp, named servers. I also have cups set up to
> > > > > > function as a print server, and sendmail is being used as a relay to our
> > > > > > mail server. yum-cron appears to be updating everything daily as
> > > > > > desired.
> > > > > >
> > > > > > This Centos 5.1 gateway was created to replace a Fedora Core 5 system
> > > > > > with the same functions. When I set up cups I copied the FC5
> > > > > > subdirectory into the Centos 5.1 system. Everything appears to work as
> > > > > > expected.
> > > > > >
> > > > > > For some reason I am having sendmail crator without giving me a notice
> > > > > > in the logs as far as I can determine. However, when I looked at the
> > > > > > messages logs I found it filled with many entries in the form of :
> > > > > >
> > > > > > kernel: ipp[24519]: segfault at 0000000000000000 rip 00002aaaabf2abc0
> > > > > > rsp 00007fff25495348 error 4
> > > > > >
> > > > > > I've done a google search and there appeared to be a problem with
> > > > > > doubled printer entries in the printer.conf causing the log entry, but
> > > > > > after checking my printer.conf file everything appears to be ok. This
> > > > > > file also worked on FC5 without a problem.
> > > > > >
> > > > > > I am not sure if the cups problem is related to the sendmail problem.
> > > > > >
> > > > > > Any ideas?
> > > > > >
> > > > > > Greg Ennis
> > > > > >
> > > > >
> > > > > I have tried several things in the past 48 hours to try to understand
> > > > > what is happening. The only thing I have done that seems to have made a
> > > > > difference is that I have switched 2.6.18-53.1.14.el5xen to
> > > > > 2.6.18-53.1.14.el5.
> > > > >
> > > > > I am still getting the following log entries :
> > > > >
> > > > > May 8 08:23:14 DeGw kernel: ipp[11677]: segfault at 0000000000000000
> > > > > rip 00002aaaabf2abc0 rsp 00007fff7b7ff6b8 error 4
> > > > > May 8 08:28:26 DeGw kernel: ipp[12925]: segfault at 0000000000000000
> > > > > rip 00002aaaabf2abc0 rsp 00007fffa5ab6988 error 4
> > > > >
> > > > > However I have only received 7 of these since last night instead of
> > > > > hundreds while running xen.
> > > > >
> > > > > When I booted the system last night with 2.6.18-53.1.14.el5 I received
> > > > > some additional "messages" log entries :
> > > > >
> > > > > May 8 01:50:00 DeGw kernel: irq 193: nobody cared (try booting with the "irqpoll" option)
> > > > > May 8 01:50:00 DeGw kernel:
> > > > > May 8 01:50:00 DeGw kernel: Call Trace:
> > > > > May 8 01:50:00 DeGw kernel: <IRQ> [<ffffffff800b703a>]__report_bad_irq+0x30/0x7d
> > > > > May 8 01:50:00 DeGw kernel: [<ffffffff800b726d>] note_interrupt+0x1e6/0x227
> > > > > May 8 01:50:00 DeGw kernel: [<ffffffff800b677f>] __do_IRQ+0xc7/0x105
> > > > > May 8 01:50:00 DeGw kernel: [<ffffffff80011cc5>] __do_softirq+0x5e/0xd5
> > > > > May 8 01:50:00 DeGw kernel: [<ffffffff8006b3bd>] do_IRQ+0xe7/0xf5
> > > > > May 8 01:50:00 DeGw kernel: [<ffffffff80069d0e>] default_idle+0x0/0x50
> > > > > May 8 01:50:00 DeGw kernel: [<ffffffff8005c615>] ret_from_intr+0x0/0xa
> > > > > May 8 01:50:00 DeGw kernel: <EOI> [<ffffffff80069d37>] default_idle+0x29/0x50
> > > > > May 8 01:50:00 DeGw kernel: [<ffffffff80046fb1>] cpu_idle+0x95/0xb8
> > > > > May 8 01:50:00 DeGw kernel: [<ffffffff803d3806>] start_kernel+0x220/0x225
> > > > > May 8 01:50:00 DeGw kernel: [<ffffffff803d3237>] _sinittext+0x237/0x23e
> > > > > May 8 01:50:00 DeGw kernel:
> > > > > May 8 01:50:00 DeGw kernel: handlers:
> > > > > May 8 01:50:00 DeGw kernel: [<ffffffff801dc154>] (usb_hcd_irq+0x0/0x55)
> > > > > May 8 01:50:00 DeGw kernel: Disabling IRQ #193
> > > > >
> > > > > If any of you can help me get a start on this problem I would sure
> > > > > appreciate your help.
> > > >
> > > > It appears to be a flakey USB controller. Is the printer USB?
> > > > Is it plugged into a USB hub? Check your USB cables, hubs and
> > > > try to plug it into a different port.
> > > >
> > >
> > > Ross,
> > >
> > > Thanks for your response. All the printers are networked and none are
> > > connected to the server's usb ports. In fact there is nothing
> > > connected to this server's usb ports.
> > >
> > > In looking at the cups error logs I am getting abundant entries of :
> > >
> > > (/usr/lib/cups/backend/ipp) crashed on signal 11!
> > >
> > > I turned on cups debug, but did not learn anything else.
> > >
> > > Thanks again for your help,
> >
> > Well the IRQ error is definitely USB related. Maybe a memory check
> > is in order here.
> >
> > Sendmail and cups are not related and the fact that you see errors
> > in both and strange runaway IRQ errors seems to mean there is a
> > hardware problem here.
> >
> > Check memory (low hanging fruit), then it may be time to start
> > looking at the motherboard.
> >
> > -Ross
> >
>
> Ross,
>
> I was hoping it might be something else. I'll check the memory tonight
>
> Thanks for your suggestions!!!
>
> Greg
> _______________________________________________
Everyone,
I finally found the solution to this problem. I did replace the memory
but there was no improvement.
When I created this machine I decided to swap the /etc/cups directory on
the old Fedora 5 system with the original Centos cups directory.
All I did to solve the problem is to use the original Centos /etc/cups
directory while manually adding the printers and remove the /etc/cups
directory that I had copied from the Fedora 5 system
This resolved the problem.
Greg
_______________________________________________
CentOS mailing list
CentOS@xxxxxxxxxx
http://lists.centos.org/mailman/listinfo/centos
