Re: [Tracker] Tracker on Solaris



On 31/01/13 16:20, Frank Lahm wrote:
Hi Martyn,

Hello,

2013/1/29 Martyn Russell <martyn lanedo com>:
On 29/01/13 14:39, Frank Lahm wrote:
The Python examples from freedesktop work just fine too... :/

Would it help running tracker-miner-fs from gdb?

I think so. The DBus error is too generic to know without that or looking at
the specific version/code I would say.

I'm using Tracker master HEAD.
dbus-daemon is 1.2.28.
libdbus-glib (if that matters): 0.88

Er, we don't use this any more. WE use gdbus as part of glib.

glib2: 2.28.6

This should be fine.

Here's the debugger step through up the AddMatch() error:

(dbx) stop at tracker-writeback-listener.c:221
(dbx) run
Running: tracker-miner-fs
(process id 16330)
Initializing tracker-miner-fs...
GLib-GIO-Message: Using the 'memory' GSettings backend.  Your settings
will not be saved or shared with other applications.
Tracker-Message: Setting up monitor for changes to config
file:'/root/.config/tracker/tracker-miner-fs.cfg'
Starting log:
   File:'/root/.local/share/tracker/tracker-miner-fs.log'
t 1 (l 1) stopped in tracker_writeback_listener_new at line 221 in
file "tracker-writeback-listener.c"
   221           miner =  g_initable_new (TRACKER_TYPE_WRITEBACK_LISTENER,
(dbx) step
t 1 (l 1) stopped in tracker_writeback_listener_get_type at line 82 in
file "tracker-writeback-listener.c"
    82   G_DEFINE_TYPE_WITH_CODE (TrackerWritebackListener,
tracker_writeback_listener, G_TYPE_OBJECT,
(dbx) step
t 1 (l 1) stopped in g_once_init_enter at line 347 in file "gthread.h"
   347     if G_LIKELY ((gpointer) g_atomic_pointer_get (value_location) != NULL)
(dbx) step
t 1 (l 1) stopped in g_once_init_enter at line 350 in file "gthread.h"
   350       return g_once_init_enter_impl (value_location);
(dbx) step
t 1 (l 1) stopped in g_once_init_enter at line 351 in file "gthread.h"
   351   }
(dbx) step
t 1 (l 1) stopped in tracker_writeback_listener_get_type at line 82 in
file "tracker-writeback-listener.c"
    82   G_DEFINE_TYPE_WITH_CODE (TrackerWritebackListener,
tracker_writeback_listener, G_TYPE_OBJECT,
(dbx) step
t 1 (l 1) stopped in tracker_writeback_listener_get_type at line 82 in
file "tracker-writeback-listener.c"
    82   G_DEFINE_TYPE_WITH_CODE (TrackerWritebackListener,
tracker_writeback_listener, G_TYPE_OBJECT,
(dbx) step

(tracker-miner-fs:16330): GLib-GIO-CRITICAL **: Error while sending
AddMatch() message: The connection is closed

t 1 (l 1) stopped in tracker_writeback_listener_new at line 227 in
file "tracker-writeback-listener.c"
   227           if (internal_error) {
(dbx) quit

Can you make any sense of that?

Thanks for trying this.

Yea, but none of this makes me think Tracker is the problem here sadly. I was hoping to find out why DBus is saying this. I did some googling and found this (among other things):

http://openbsd.7691.n7.nabble.com/gnome-startup-error-td220228.html

It's the same problem you mention but not just Tracker, other services too.

It looks to me like we're trying to use the connection after it's been dropped. A lot of the Google results seem to show this error AFTER a session (lightdm for example) is killed or perhaps even the dbus daemon itself. With your logs, it's just dropping the connection but there are no signs of any other foul play. I wonder why it drops the connection.

It looks like it's either a permissions/configuration issue with DBus on Solaris OR an old version which isn't so compatible to gdbus? The Python examples I sent you use the old school dbus approaches I believe.

I found some newer gdbus examples in python too, might be worth trying:

  Server:
  http://people.canonical.com/~pitti/tmp/gdbus-server.py

  Client:
  http://zurlinux.com/?p=673

Still, that will really just tell you if this is a compatibility issue, it doesn't fix the problem and you have a recent GLib too - so :/

--
Regards,
Martyn

Founder and CEO of Lanedo GmbH.



[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]