Hi Team. I’ve been using TVHeadend to drive both the front and backend server on a single RPi 2. Tuner is an HDHomeRun Connect. HDHomeRun is connected directly to the Pi through the ethernet port, using external USB hard drive for live TV buffer/storage. Wifi network is bridged to provide connectivity to the system. Everything has been working great, except I’m getting random lock ups. Live TV will be display normally for 5 minutes to 2 hours (occurs randomly), and then the audio and video freezes. The system becomes unresponsive to CEC and Kodi App Remotes, unable to SSH in at this point. Strangely, the network activity lights continue to blink on their merry way and the systems appears to be running, except I cannot access it without a hard reboot.
Kodi log from most recently lockup: http://xbmclogs.com/pa0gbsf2j. System locked up at approx. 16:50, no immediate sign of issue.
I don’t think it is a power supply issue, all accessories are running through a powered hub and the pi is on a separate dedicated 2A supply. No overclock on the system. I’m thinking I may be maxing out the USB bandwidth of the device and creating the lock up situation, but that’s just a guess at this point. Anyone run into something similar?
No sign of the coloured square in top-right corner of screen?
As soon as you notice the video freeze, can you confirm if you can ssh in first. Assuming not, do you get any response from remote (e.g. bring up OSD)?
Any different with omxplayer disabled (but MMAL enabled)?
No sign of the rainbow power icon in the corner. I believe the power supply is sufficient, but it is possible it’s sagging under load.
As soon as it freezes, no response to SSH, phone remote, or CEC remote.
Made the switch to MMAL, and had over an hour of no issues. Finally received the sad face screen and OSMC soft booted back to the main screen. Different error than before, and not the complete lockup as seen previously. Testing again to see if I can reproduce. Here’s the relevant log entries around the failure:
Jan 03 16:51:59 osmc tvheadend[683]: htsp: 127.0.0.1 [ osmc | Kodi Media Center ]: Disconnected
Jan 03 16:51:59 osmc tvheadend[683]: subscription: 0002: “127.0.0.1 [ osmc | Kodi Media Center ]” unsubscribing from “10.1 KSAZ”, hostname=“127.0.0.1”, username=“osmc”, client=“Kodi Media Center”
Jan 03 16:51:59 osmc tvheadend[683]: htsp: 127.0.0.1 [ osmc | Kodi Media Center ]: Write error – Broken pipe
Jan 03 16:51:59 osmc sudo[407]: pam_unix(sudo:session): session closed for user osmc
Jan 03 16:51:59 osmc mediacenter[370]: Kodi exited with return code 139 after 1 hours, 1 minutes and 9 seconds
Jan 03 16:51:59 osmc kernel: Console: switching to colour frame buffer device 240x67
Jan 03 16:52:10 osmc mediacenter[370]: Starting Kodi…
Experienced a full lock up with MMAL this time. Regular Kodi logs not showing anything out of place other than some CDVDPlayerAudio::HandleSyncError messages.
Is there a way to recover the OSMC log from the last boot?
kodi.old.log from the same location.
Thanks - found it. Again, nothing out of the ordinary. No additional data is written as soon as the system locks up, so I think I’m missing anything that would help narrow down the failure.
Starting to think a full system rebuild from scratch may be in order unless I can narrow it down further. Hoping to avoid that if at all possible.
A second sdcard would be useful. Otherwise you can backup the current sdcard (e.g. with win32diskimager) and thne try a clean install. With the clean install just make the minimum changes needed to get live tv working (don’t install lots of skins and add-ons and leave as many settings as you can on default). If that seems better then you can gradually install add-ons and change settings, testing after each change.
If it doesn’t help then you can always revert the sdcard to the backup.
Narrowed it down to either my Bluetooth implementation or an out of control add-on. With those disabled I haven’t had a lock up since. Will see if I can narrow it down to a specific factor. Thanks for the help!
Did you ever determine which add-on was causing the issue?