Samba: CUPS printers are not available at startup

[Lucio M Nicolosi]

On Sun, May 1, 2011 at 4:03 PM, NoOp <glgxg at sbcglobal.net> wrote:

> On 04/22/2011 08:36 PM, Lucio M Nicolosi wrote:
> ...
> >
> > As intended, stripped down my hardware and thoroughly tested the
> > behavior of the smbd daemon. Removed Firestarter and iptables are now
> > as naked as a newborn. USB ports at use:  the printers, a mouse and a
> > camera (forgot to remove it but it is more recent than the problem).
> >
> > Conclusion: in my station it doesn't work, neither on Maverick 64 or
> > Natty 64. Something blocks the daemon's auto reload function. So the
> > parameter "printcap cache time ="  is useless. Note that the Natty
> > install is bare, almost nothing but the default itens.
> >
> > On a headless Lucid 64 station nearby, the adding or removing of a
> > printer from Cups server is almost immediately reflected by the
> > command sbmclient -L [IP], indicating the daemon is working.
> >
> > A typical segment of the Maverick 64 /var/log/samba/smbd log file at
> > the faulty station is:
> >
> > [begin]
> >
> > [2011/04/22 12:53:21,  0] smbd/server.c:1123(main)  smbd version 3.5.4
> started.
> >   Copyright Andrew Tridgell and the Samba Team 1992-2010
> > [2011/04/22 12:53:21.406311,  0]
> > printing/print_cups.c:108(cups_connect)  Unable to connect to CUPS
> > server localhost:631 - Connection refused
> > [2011/04/22 12:53:21.411729,  0]
> > printing/print_cups.c:108(cups_connect)  Unable to connect to CUPS
> > server localhost:631 - Connection refused
> >
> > {(above) normal, denotes cups starting after smbd, a known bug}
> >
> > [2011/04/22 12:53:21.412194,  0] smbd/server.c:1169(main)  standard
> > input is not a socket, assuming -D option
> >
> > {standard, also appears at the functional station}
> >
> > [2011/04/22 12:53:21.418346,  0]
> > smbd/server.c:500(smbd_open_one_socket) smbd_open_once_socket:
> > open_socket_in: Address already in use
> > [2011/04/22 12:53:21.418455,  0]
> > smbd/server.c:500(smbd_open_one_socket) mbd_open_once_socket:
> > open_socket_in: Address already in use
> > [2011/04/22 12:53:21.419338,  0] smbd/server.c:281(remove_child_pid)
> > Could not find child 1146 -- ignoring
> >
> > {these previous lines probably contain the clue, they are also present
> > at Natty's log, it looks like smbd is unable to open a needed socket
> > for the service daemon - ran socklist and checked /etc/init/smbd.conf
> > but haven't a clue about what socket it might be, perhaps the author,
> > Steve Langasek <steve.langasek at ubuntu.com> knows what is all about}
> >
> > [2011/04/22 12:53:27.158123,  1] smbd/process.c:776(smbd_sig_hup_handler)
> >   Reloading services after SIGHUP
> > [2011/04/22 12:53:27.171400,  0] smbd/server.c:281(remove_child_pid)
> > Could not find child 2170 -- ignoring
> > [2011/04/22 13:12:40,  0] smbd/server.c:1123(main)
> >
> > {displayed after forcing a smdb reload - this SIGHUP should run at
> > intervals set by  "printcap cache time =" }
> >
> > [end]
> >
> > After this I gave up and decided to apply the simple transitory patch
> > recommended at
> https://bugs.launchpad.net/ubuntu/+source/cups/+bug/494141,
> >  added '/sbin/reload smbd' at the end of /etc/rc.local. Although the
> > daemon still doesn't work, at least the Cups printers are loaded just
> > after startup.
> >
> > So, now my rc.local now looks like this:
> >
> >  #!/bin/sh -e
> >  # [snip]
> >  /sbin/reload smbd
> >  exit 0
> >
> > L.
> >
>
> I wonder if this is related:
> https://bugs.launchpad.net/ubuntu/+source/udev/+bug/691027
> I've noticed that after cold boot, my printers are not enabled & I have
> to reenable.
>

They are not enabled at startup, it's a known bug, Cups starts after the
Samba modules, but they should be auto enabled after an interval defined by
"printcap cache time =" parameter that defaults to 750 seconds, otherwise
you would be facing the same bug I am.


Not sure if this is related, but I'm also seeing:
> [2011/05/01 10:05:22.489679,  0] smbd/server.c:281(remove_child_pid)
>  Could not find child 3244 -- ignoring
>

It probably has something to do with "my" bug.

-- 
Lucio M. Nicolosi, Eng.
Open Source Implementation
System and Applications
GNU/Linux
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.ubuntu.com/archives/ubuntu-users/attachments/20110501/4fa379b0/attachment.html>