[chrisism]

Same issue here. My headless crashes when its calling the addons.xml stuff. Anybody a good solution? My guisettings.xml gets overwritten each time I start my docker container.

[sgeoxd]

As a workaround I got it to stop crashing by copying Addons20.db from my 16.1 Mac OS X box and set the below to 2 in guisettings.xml to prevent addon updates:

Code:

<addonupdates>2</addonupdates>

Not pretty, but for now it'll do.

Had issues with the Linuxserver.io docker container constantly crashing. Thanks so much for this, works like a charm!

[quixers]

As a workaround I got it to stop crashing by copying Addons20.db from my 16.1 Mac OS X box and set the below to 2 in guisettings.xml to prevent addon updates:

Code:

<addonupdates>2</addonupdates>

Not pretty, but for now it'll do.

Awesome fixed my headless kodi thanks.

[chrisism]

Mmm.. Mine is still crashing but it looks like my guisettings xml gets overwritten each time. I am using the celedhrim version. Where do I have to change the guisettings.xml actually

[joelones]

Mmm.. Mine is still crashing but it looks like my guisettings xml gets overwritten each time. I am using the celedhrim version. Where do I have to change the guisettings.xml actually

Are you asking where is guisettings.xml located? If so, it's located in the userdata folder. The <addonupdates> tag is within the <general> tag. And lastly, guisettings.xml gets overwritten because you probably haven't stopped the container...

[chrisism]

Mmm.. Mine is still crashing but it looks like my guisettings xml gets overwritten each time. I am using the celedhrim version. Where do I have to change the guisettings.xml actually

Are you asking where is guisettings.xml located? If so, it's located in the userdata folder. The <addonupdates> tag is within the <general> tag. And lastly, guisettings.xml gets overwritten because you probably haven't stopped the container...

Yeah its a bit strange here. I now changed the guisettings.xml again in my userdata folder and started the container and now its running again. Last week I got it running and when I was looking at it today the container was stopped and the guisettings again had the 0 parameter for addonupdates.
I didnt restart the container or linux host its running in.

Also before I changed the param each time and after I started the container the guisettings xml had the old value again. Just was wondering if there is something that is overwritting the settings on start of the container?
At least its running again now. Still, would be nice if we have a proper fix for it indeed.

[fritzl]

I had the same problem. Turns out you have to remove the `default="true"` part, i.e. change <addonupdates default="true">0</addonupdates> to <addonupdates>2</addonupdates>. At least that worked for me

[chrisism]

I had the same problem. Turns out you have to remove the `default="true"` part, i.e. change <addonupdates default="true">0</addonupdates> to <addonupdates>2</addonupdates>. At least that worked for me

That seems to be it indeed! Thanks

[matthuisman]

I had same segfault issue as well using the Linuxserver.io on an Odroid XU4.
The <addonupdates>2</addonupdates> fixed it for me

Anyone know how to disable the
CAESinkOSS::EnumerateDevicesEx - No OSS mixer device present: /dev/mixer

It just takes quite a while it keeps retrying.

New issue
When trying to view any media via the web interface, it crashes.
Normal JSON requests seem fine, but straight after the below last debug line, KODI is crashing and restarting..

08:18:21 T:3022482680 DEBUG: webserver: request received for /jsonrpc?Player.GetActivePlayers
08:18:21 T:3022482680 DEBUG: JSONRPC: Incoming request: {"jsonrpc":"2.0","method":"Player.GetActivePlayers","id":1}
08:18:22 T:3022482680 DEBUG: webserver: request received for /jsonrpc?VideoLibrary.GetTVShows
08:18:22 T:3022482680 DEBUG: JSONRPC: Incoming request: {"jsonrpc":"2.0","method":"VideoLibrary.GetTVShows","id":1,"params":{"properties":["genre","plot","title","lastplayed","episode","year","playcount","rating","thumbnail","studio","mpaa","premiered"]}}
08:18:22 T:3022482680 DEBUG: RunQuery took 12 ms for 1 items query: SELECT * FROM tvshow_view

Does it with local and MySQL databases.
Update library, profiles commands all work fine.

gdb output

Starting program: /usr/lib/kodi/kodi.bin --headless

Thread 9 "kodi.bin" received signal SIGSEGV, Segmentation fault.
[Switching to LWP 1389]
0x7fa1e148 in std::_Rb_tree_iterator<std::pair<Field const, CVariant> > std::_Rb_tree<Field, std::pair<Field const, CVariant>, std::_Select1st<std::pair<Field const, CVariant> >, std::less<Field>, std::allocator<std::pair<Field const, CVariant> > >::_M_emplace_hint_unique<std::piecewise_construct_t const&, std::tuple<Field&&>, std::tuple<> >(std::_Rb_tree_const_iterator<std::pair<Field const, CVariant> >, std::piecewise_construct_t const&, std::tuple<Field&&>&&, std::tuple<>&&) ()

[Tad_Dancer]

Edit:

Alright, never mind, I just forgot to enable it in the guisettings.xml again, after I updated the configuration. My bad.


Hi guys,

I've had Kodi running as a headless server for a while now, for the last year or so I ran the docker. Everything always worked great until a few days ago I had the same segfaults and boot loops.
I was able to fix it with that addon update disable. Now it's coming back up, but it doesn't start the web server for some reason and I can't figure out why.

These are the last lines of the debug log, after that nothing else is happening with the machine:

Code:

15:53:33 T:139731409016576  NOTICE: ES: Starting UDP Event server on 127.0.0.1:9777
15:53:33 T:139731409016576  NOTICE: UDP: Listening on port 9777
15:53:33 T:139731459372800   DEBUG: CPythonInvoker(0, /opt/kodi-server/share/kodi/addons/service.xbmc.versioncheck/service.py): start processing
15:53:33 T:139731459372800   DEBUG: -->Python Interpreter Initialized<--
15:53:33 T:139731459372800   DEBUG: CPythonInvoker(0, /opt/kodi-server/share/kodi/addons/service.xbmc.versioncheck/service.py): the source file to load is "/opt/kodi-server/share/kodi/addons/service.xbmc.versioncheck/service.py"
15:53:33 T:139731459372800   DEBUG: CPythonInvoker(0, /opt/kodi-server/share/kodi/addons/service.xbmc.versioncheck/service.py): setting the Python path to /opt/kodi-server/share/kodi/addons/service.xbmc.versioncheck:/usr/lib/python2.7:/usr/lib/python2.7/plat-x86_64-linux-gnu:/usr/lib/python2.7/lib-tk:/usr/lib/python2.7/lib-old:/usr/lib/python2.7/lib-dynload:/usr/local/lib/python2.7/dist-packages:/usr/lib/python2.7/dist-packages
15:53:33 T:139731459372800   DEBUG: CPythonInvoker(0, /opt/kodi-server/share/kodi/addons/service.xbmc.versioncheck/service.py): entering source directory /opt/kodi-server/share/kodi/addons/service.xbmc.versioncheck
15:53:33 T:139731459372800   DEBUG: CPythonInvoker(0, /opt/kodi-server/share/kodi/addons/service.xbmc.versioncheck/service.py): instantiating addon using automatically obtained id of "service.xbmc.versioncheck" dependent on version 2.1.0 of the xbmc.python api
15:53:33 T:139731459372800   DEBUG: Version Check: Version 0.3.17 started
15:53:38 T:139731459372800   DEBUG: Version Check: Version installed {u'major': 16, u'tag': u'releasecandidate', u'tagversion': u'2', u'minor': 1, u'revision': u'2016-03-29-be20e8a-dirty'}
15:53:38 T:139731459372800    INFO: CPythonInvoker(0, /opt/kodi-server/share/kodi/addons/service.xbmc.versioncheck/service.py): script successfully run
15:53:38 T:139731459372800    INFO: Python script stopped
15:53:38 T:139731459372800   DEBUG: Thread LanguageInvoker 139731459372800 terminating
15:53:48 T:139732323063936    INFO: Python, unloading python shared library because no scripts are running anymore
15:53:48 T:139732323063936   DEBUG: UnloadExtensionLibs, clearing python extension libraries
15:54:43 T:139731684325120   DEBUG: Thread AESinkNull 139731684325120 terminating

Any ideas how I can get the machine to start the webserver? Does anyone else have similar problems?

Edit:
Oh, and if that matters, the necessary plugins are getting initialized at startup:

Code:

15:53:33 T:139732323063936   DEBUG: ADDON: cpluff: 'Plug-in webinterface.default has been installed.'
15:53:33 T:139732323063936   DEBUG: ADDON: cpluff: 'Plug-in xbmc.webinterface has been installed.'

But there is nothing in the Logs about even an attempt to start it.

[quixers]

Has anyone tried the headless patches with Kodi v17 Krypton?

Not far off release I think..

[celedhrim]

Has anyone tried the headless patches with Kodi v17 Krypton?

Not far off release I think..

Yes , it works since kodi 17 alpha3

https://github.com/Celedhrim/docker-kodi...ts/krypton

[TopperBG]

Sorry for lame q, but is supposed to have fully installed Kodi (I have Jarvis @LibreElec) and to copy my install from /storage/.kodi to dedicated Debian server in for exmpl /kodi-headless ?
And after that run docker container for Jarvis ?
Thanks in advance for any response

[thezoggy]

just to note, its been 8 days since I rebooted my nas after upgrading unraid (6.1.9->6.2.2).. which caused my dockers to get rebuilt.. since then I've yet to see kodi-headless segfault. thus no workaround/hack has been done.. so looks like the issue magically went away.. as both kodi and linuxserver said they havent changed anything. so maybe there was just a docker bug?

[quixers]

Has anyone tried the headless patches with Kodi v17 Krypton?

Not far off release I think..

Yes , it works since kodi 17 alpha3

https://github.com/Celedhrim/docker-kodi...ts/krypton

Awesome thanks for keeping this updated. I'm using it outside docker, used the docker script as a guide. Its working fine on b3. Tried patch on b5 but didn't seem to work. For library updates I'm sure b3 is fine..