3B+ freezing up?

I’m luck enough to have a couple of new 3B+, which I’ve swapped in place of my previous 3B using the same SD card. Over a few days of testing (i.e. watching movies) I’ve had both freeze playing video, requiring a power cycle. No SSH access at this point.

I’d also installed new licences for MPG2 and VC1 codecs (as per previous on the 3B).

Is anyone else seeing similar please?

Best regards,
Mark W.

Did you get an appropriate power supply?

I’m running them both with nearly new Raspberry Pi original 2.5A supplies, which I beleive are still acceptable for the B+?

Well than I suggest you enable debug logging and upload logs so that we might be able to spot something.

I’m seeing a similar issue on a Pi 3 B+ - I’ve narrowed it down to the PVR WMC client at least for me. If I don’t have that add-on enabled, OSMC doesn’t crash. As soon as I enable it, it will freeze first while importing guide from clients, and after pulling the power it will come back and run a bit longer, sometimes letting me watch video.

Logs would definitely help. I will say that I was having similar problems, so I moved to a less aggressive (and smaller) cache for video and my problems went away.

Specifically I had this:

    <cache>
        <memorysize>157286400</memorysize>
        <buffermode>1</buffermode>
        <readfactor>20</readfactor>
    </cache>

and changed to this:

    <cache>
        <memorysize>104857600</memorysize>
        <buffermode>1</buffermode>
        <readfactor>12</readfactor>
    </cache>

I’m working under the theory that with the 3B the 100mb ethernet was helping to slow things down enough that these settings didn’t cause problems, but with the 3B+ the 300mb ethernet I get now was.

Again though, logs would help, as your issue could be totally unrelated to mine.

In typical fashion, since asking the question neither of my 3B+ have frozen again.

I did initially have one in case (now removed) , though I suspect that’s a red herring as the CPU temperatures looked good up until the freeze.

Are you using wired ethernet? Sam was muttering something about too many interrupts as if the driver needs some more work. A read factor of 20 (@pkscout) could be stressing it too far. Even 12 looks a lot.

I am using the wired ethernet port. Since moving it to a readfactor of 12 I haven’t had any problems, but I guess I could ease it down some more and see if I get any stuttering video. Unrelated, but I wonder if the ethernet interrupt issue could be contributing to the problem I have with my GPIO IR sensor. It works fine in menus and even with up to 720P video, but at 1080P the remote becomes mostly unresponsive (irw doesn’t even show the codes being received). All the video is streamed from an NFS share on another machine. The GPIO IR was a new project, so I just shelved it for now and went back to my FLIRC, which has been rock solid.