Clients getting "kicked" with "tuner no longer available" after a time

[smajor]

I reloaded 3.1.1 from scratch which got rid of some sever client problems I was having. The only one left is, when a 2nd client starts a stream, sooner or later it will be kick with a "Tuner no longer available" message. It doesn't matter which device is in use, it is always the 2nd client that is started that gets booted.

Obviously, for me, this is new in 3.1.1 - 2.6.2 was rock solid. Hours and hours of streaming live TV with two HD-PVRs in use with no problems.

Logs attached. The 2nd tuner was in kicked while watching "All About Dogs" for reference.

[sub]

What client were you using exactly? Was this XBMC?

You'll need to zip and attach the complete set of logs, so I can see what's happening in nrecord.log etc.

[smajor]

Try these. It's from this morning.

The logs would show it from XBMCs plugin. It really doesn't matter which, as it will happen with XNEWA or the "Touch" interface as well.

It's always the 2nd stream that is started that will get kicked (I don't have a 3rd device to test.) Recordings with both tuners simultaneously work just as well as ever.

[sub]

Do they both run for a period of time, or does the second one die almost immediately?

[sub]

There is a lot of log there. I don't suppose you remember exactly what time the error happened? Nothing is standing out, but I "cant see the forest for the trees".

[smajor]

Around 9am is the first HD-PVR tuning to channel 1884.

Then look around 9:42am when it tunes to channel 123. That's the 2nd HD-PVR going into live TV mode.

Everything is fine until about 9:48am (I think) when it stops.

[sub]

Your web.log only go back as far as 10:30am, so that might explain why I'm not seeing the things I'm looking for.

Sounds like you'll need to reproduce it again, and repost the logs. Take notes of the exact time the problem happens if you can.

[smajor]

Sure, it's easy to reproduce. Attached.

Channel 1884 is the first tuner.

Channel 109 is the second and quits about 6-7 minutes in. Right around 4:20pm.

[sub]

It looks like the second file stopped growing, and after 8 seconds of waiting for it to grow, it eventually gave up and decided it had reached the end of the stream. I'm not sure why it would have stopped growing though.

There is some extra logging that can be enabled to show us the data coming from the device, but unfortunately it'll only log the first device, and your problem is happening on the second one.