Roku Developer Program

[destruk]

SetMaxVideoDecodeResolution might help with the ram --
http://sdkdocs.roku.com/display/sdkdoc/ ... egerasVoid

[greubel]

Gave SetMaxVideoDecodeResolution() a try and video failed after 0:03 minutes.
Thought I should try again (too short) and it ran for 1:41 minutes before it took a dive.

Would it be possible for someone to look at one of my Roku dumps ?

[RokuDale]

Which platforms does this happen on? Roku 2? Roku 3? Roku TV? Roku 1?
All of the above?

[greubel]

I tried it on all my platforms, same movie, took awhile.
Roku(3.1 1198) Model N1000
Roku(3.1 1198) Model 2100X
Roku(5.5 453) Model 3100X <-------------- fails
Roku(5.5 415) Model 4200X

Being connected by Wifi (12-20 mbps) or Lan(40+ mbps) doesn't seem to make a difference. Before it crashes, things start taking a long time. Looking at my traces entries, things that normally take 10 to 20 ms start taking close to a second. Sometimes in my message loop, I will get 2 or 3 video position events queued at the same time. The video position events are set at 1 per second. That says that I didn't get control back to check the queues for quite awhile. So something is hogging the CPU or it's just out of memory. I'ld say memory, since the next thing it does is REBOOT.

The only thing that squeaks it by, is for me to limit the bandwidth coming out of my server to below 10 mbps. Checking the BANDWIDTH log entry, the Roku says it's around 8 mbps. But that isn't a good fix for all the other DLNA media servers.

SURE would be nice if there was a method that returned in use and available memory !

[RokuDale]

I don't believe it has anything to do with available memory.

[greubel]

In a previous thread I sent a movie to Joel. It always failed across the LAN, Wifi but played successfully if pulled from the web.

I'm at a loss. Users have complained to me and on other forums about a similar problem. Movie plays and half way you pause it and the remote is unresponsive OR screen becomes real slow and jerky and then it crashes. Sure would be nice to have some sort of resolution.

I guess I could try putting in some code to steal bandwidth from the video player to see if that has any effect.

[TheEndless]

Hey greubel, any chance you have multiple roSystemLogs created? I ask, because I once saw similar behavior (albeit after a much longer period of time), and it turned out that the message queues in the roSystemLogs that weren't being monitored were filling up and bogging down the box (at the time, I was erroneously creating a new roSystemLog for each screen, so all of the roSystemLogs associated with the screens lower in the stack were filling up). Higher bandwidth could result in more http connects, filling up the message queues faster.

[RokuDale]

Is the movie you sent to Joel the same one you used for testing on all of those platforms?
I presume also that there is no transcoding being done. You are just directplaying it?
The same problem should occur using RMP.
dale

[greubel]

It occurs with just about every movie. There is one roSystemLogs created at app init.
I need to try RMP again. I think I tried at the beginning of this fiasco but need to make sure it fails.
Get back to you !

[RokuDale]

Does it still happen if you disable the creation of the roSystemLog?
That would remove that as a memory hog.
dale