I also experienced strange network device breakdowns on the Wetek Play 2 since many LE versions. Today I started to hunt that and maybe my log will help the LE developers to catch this nasty bug:
My LE version is compiled from github master and has version: devel-20170106171602-r25136-g3adb69f 8.0, kernel: Linux ARM 64-bit version 3.14.29 (installed to NAND). All services are disabled: no Zeroconf, no Webserver, no UPnP / DLNA, no Airplay. Skin is Confluence. I use TVheadend 4.1.2309 with ?scam and this runs quite well - with the known bugs described here in the forum.
But always when I watch a longer movie from my NAS or stream Live TV to a clients VLC, the network goes down. So I switched on Debug-Logging with all components and asked VLC to view an HD broadcasting from the wetek server to my laptop (connected via WLAN to my homenet; wetek is connected to the LAN device). And within some few minutes, the LAN device broke down (it doesn't matter whether it is DHCP or static configured).
Here is my log: Ubuntu Pastebin
I started my Wetek Play 2 at 2:55, waited 5 minutes and started VLC at 3:00. At 3:02 the eth0 device collapsed. What I can see in the log: there is a lot of UPnP talkings to my router and my NAS (syno DS412+), although I switched that off! I consider UPnP as insecure and avoid to use it anyway. I haven't thought that UPnP service is running when it is disabled ;-P
After 03:02:49.235 T:545996714208 INFO: Platinum [neptune.http]: cleaning up connection (1 remain) - I just waited some seconds and swithed my box off and booted it to SD-card to save the log. Why is connmand not able to bring the device up again?
Hope my investigation will help
ajb
not a solution, plus an issue for highbit rate 4k files i'm afraid.
