I experience this every boot. Sooner or later I notice my CPU is at 70 degrees Celcius and find out that one of my 8 cores is at 100% utilization. I suspect that some photo/video data stored on my notebook causes this. Can you suggest what additional logging I need to enable to catch it? I can't see anything related to gnome-photos in the logs now. I can't kill the gnome-photos process with any signal. I can't shutdown/reboot my machine while this hanging condition is prresent. I have to force poweroff.
How do you suspect that? With "top" output? What error do you get when using "killall -9 gnome-photos"?
(In reply to Pacho Ramos from comment #1) > How do you suspect that? With "top" output? What error do you get when using > "killall -9 gnome-photos"? I have a gnome extension (Freon) to show my CPU temp in the top bar. The information comes from lm_sensors. I check the CPU utilization with "top". I use "kill -9" with gnome-photos pid to terminate gnome-photos, however I don't have any error message back. kill returns w/o any message, but the gnome-photos process continues to run.
Maybe running gnome-photos from a terminal will show any error or warning that could give us a hint about what is causing it to eat resources without limit
in fact, I've never run gnome-photos myself before :-( it was started by some background process. When I started it from a terminal, it produces a lot of messages like this: (gnome-photos:18535): Tracker-CRITICAL **: tracker_sparql_cursor_real_get_double: assertion '_tmp1_ == TRACKER_SPARQL_VALUE_TYPE_DOUBLE' failed ** (gnome-photos:18535): WARNING **: Unable to create thumbnail: ?????? HTTP: Internal Server Error (gnome-photos:18535): Tracker-CRITICAL **: tracker_sparql_cursor_real_get_integer: assertion '_tmp1_ == TRACKER_SPARQL_VALUE_TYPE_INTEGER' failed ** (gnome-photos:18535): WARNING **: Unable to query info for file at file:///home/anton/%D0%97%D0%B0%D0%B3%D1%80%D1%83%D0%B7%D0%BA%D0%B8/Igo_2.4_9.6.29.468978/iGO/themechanger/meteo/day/16.png: Error when getting information for file '/home/anton/????????/Igo_2.4_9.6.29.468978/iGO/themechanger/meteo/day/16.png': No such file or directory however, I couldn't reproduce the problem this way. I should say however that moving through my google photo archives loads CPU a lot, but it goes back to idle w/o problem. Maybe my 100% cpu effect is caused by an attempt to process my google image archive?
Maybe unrelated but I have seen rygel take 100% (or more) CPU while being unresponsive to network requests on launch, when tracking is indexing. This also caused a slow down of tracker indexing resulting in very long start time. I suggest your reset your tracker database, due to error messages you pasted, then let it index your system again and try to start gnome-photos once it has finished. If you have folders you do not wish to index (build/source folders), you can tweak tracker using tracker-preferences.
Did you try the above tip?
Yes, I did. It didn't help me at all unfortunately. I ended up removing gnome-photos from my system to avoid this problem.
Strange, the problem is that without being able to reproduce on our side, it will be hard to fix it or to report to upstream to try to go deeper (bugzilla.gnome.org)
Are you still suffering this with 3.20?
3.20 seems to be fine
thanks for feedback :)
