OpenDCT: An open source digital cable tuner.

[EnterNoEscape]

Originally Posted by reggie14

Would there be any advantage to using OpenDCT with my HDHR Connect? (And does that even work?)

For background, I've noticed major AV sync problems with Plex on my HDHR Connect recordings, while Plex does a lot better with my HDHR Prime recordings. The main difference is that the HDHR Prime records transport streams through OpenDCT, while the HDHR Connect records program streams through Sage.

OpenDCT supports that model and should work fine with it. I use one for testing ATSC and it should be fine for ClearQAM too. The biggest advantage that I've seen people experiencing is being able to share their tuners with other programs. You can also have SageTV record transport streams instead of program streams if you want to test that route out first.

[tmiranda]

OpenDCT stopped working again. Log attached.

[tmiranda]

Here are some older logfiles as well.

[EnterNoEscape]

Originally Posted by tmiranda

OpenDCT stopped working again. Log attached.

This problem is looks very different from the first one you had. Can you download from http://192.168.1.25:5004/auto/v55 in a browser? OpenDCT couldn't connect to that URL, then it switched over to RTP and tuned the channel in via the native method. After that was done, very little or no data at all was actually streaming from the HDHomeRun. This usually happens when a channel is copy protected in some way but I don't see any evidence that this was the case, so it feels more like a network issue or device issue. I'm leaning more on it's a device issue based on the line readout from vstatus from the HDHomeRun itself which contains PACKETS_PER_SECOND=0 even with RTP showing that it's active and configured on the exact same line. streaminfo was returning 1: 64 ANIMAL which further supports channel 55 not actually existing on the frequency the CableCARD was mapping channel 55 to.

I did spot a parsing error that I didn't know was possible and will be fixed in the next release, but it has to do with parsing some info for debugging and doesn't have any effect on tuning.

Much later I can see that you start to get the OutOfMemory errors again, getting them without any caching makes me think there might be a memory leak related to the FFmpeg cleanup and deallocation process when the channel is not coming in. I can see in your log that this happens hundreds of times for that one channel.

Edit: The archived logs from the past few days show very normal activity.

[tmiranda]

Originally Posted by EnterNoEscape

This problem is looks very different from the first one you had. Can you download from http://192.168.1.25:5004/auto/v55 in a browser? OpenDCT couldn't connect to that URL, then it switched over to RTP and tuned the channel in via the native method. After that was done, very little or no data at all was actually streaming from the HDHomeRun. This usually happens when a channel is copy protected in some way but I don't see any evidence that this was the case, so it feels more like a network issue or device issue. I'm leaning more on it's a device issue based on the line readout from vstatus from the HDHomeRun itself which contains PACKETS_PER_SECOND=0 even with RTP showing that it's active and configured on the exact same line. streaminfo was returning 1: 64 ANIMAL which further supports channel 55 not actually existing on the frequency the CableCARD was mapping channel 55 to.

I did spot a parsing error that I didn't know was possible and will be fixed in the next release, but it has to do with parsing some info for debugging and doesn't have any effect on tuning.

Much later I can see that you start to get the OutOfMemory errors again, getting them without any caching makes me think there might be a memory leak related to the FFmpeg cleanup and deallocation process when the channel is not coming in. I can see in your log that this happens hundreds of times for that one channel.

Edit: The archived logs from the past few days show very normal activity.

One odd thing I noticed was the Sage service was running under LocalSystem instead of the user I had assigned. This may have to do with my upgrading Sage. Would that cause those weird errors? I changed the service back to run under the user.

[tmiranda]

Originally Posted by EnterNoEscape

This problem is looks very different from the first one you had. Can you download from http://192.168.1.25:5004/auto/v55 in a browser? OpenDCT couldn't connect to that URL, then it switched over to RTP and tuned the channel in via the native method. After that was done, very little or no data at all was actually streaming from the HDHomeRun. This usually happens when a channel is copy protected in some way but I don't see any evidence that this was the case, so it feels more like a network issue or device issue. I'm leaning more on it's a device issue based on the line readout from vstatus from the HDHomeRun itself which contains PACKETS_PER_SECOND=0 even with RTP showing that it's active and configured on the exact same line. streaminfo was returning 1: 64 ANIMAL which further supports channel 55 not actually existing on the frequency the CableCARD was mapping channel 55 to.

Forgot to mention, browser window is empty when I try to connect to that URL. Is my prime dead? I power cycled it and that made no difference.

[UgaData]

Originally Posted by tmiranda

One odd thing I noticed was the Sage service was running under LocalSystem instead of the user I had assigned. This may have to do with my upgrading Sage. Would that cause those weird errors? I changed the service back to run under the user.

I wouldn't think this to be an issue, unless your recording directory/folder is on Networked storage that needs the LAN user rights to be accessed.

[EnterNoEscape]

Originally Posted by tmiranda

Forgot to mention, browser window is empty when I try to connect to that URL. Is my prime dead? I power cycled it and that made no difference.

I did not necessarily mean the hardware is completely broken so much as it's providing a URL that doesn't actually work. If you go to the web interface for the device, do you have a channel 55?

[EnterNoEscape]

Originally Posted by tmiranda

One odd thing I noticed was the Sage service was running under LocalSystem instead of the user I had assigned. This may have to do with my upgrading Sage. Would that cause those weird errors? I changed the service back to run under the user.

I can see your recordings are local and not to network shares, so the local system account isn't a problem at least for recordings.

[kruzerman]

Now that i finally have OpenDCT up and running, my sage server has been rescued from the grave. All we had been using it for was to let my daughter watch all her saved shows. I have DirecTV with DVR feature but it pales in comparison to Sage. Next is to hook up the HDPVR to see if i can get DirecTV recordings in sage.

Cheers!

[tmiranda]

Originally Posted by EnterNoEscape

This problem is looks very different from the first one you had. Can you download from http://192.168.1.25:5004/auto/v55 in a browser? OpenDCT couldn't connect to that URL, then it switched over to RTP and tuned the channel in via the native method. After that was done, very little or no data at all was actually streaming from the HDHomeRun. This usually happens when a channel is copy protected in some way but I don't see any evidence that this was the case, so it feels more like a network issue or device issue. I'm leaning more on it's a device issue based on the line readout from vstatus from the HDHomeRun itself which contains PACKETS_PER_SECOND=0 even with RTP showing that it's active and configured on the exact same line. streaminfo was returning 1: 64 ANIMAL which further supports channel 55 not actually existing on the frequency the CableCARD was mapping channel 55 to.

I did spot a parsing error that I didn't know was possible and will be fixed in the next release, but it has to do with parsing some info for debugging and doesn't have any effect on tuning.

Much later I can see that you start to get the OutOfMemory errors again, getting them without any caching makes me think there might be a memory leak related to the FFmpeg cleanup and deallocation process when the channel is not coming in. I can see in your log that this happens hundreds of times for that one channel.

Edit: The archived logs from the past few days show very normal activity.

I rebooted everything (server, all HD300's, all the networking gear) and things went back to functioning properly. I'll let you know if I problems arise again. The pattern I'm seeing is that everything works great for a few days and then problems start popping up. If there is anything I can do to help you debug this let me know, I'd like to get this rock solid.

Some of my channels have a weak signal which may be a contributing factor.

[EnterNoEscape]

Originally Posted by tmiranda

I rebooted everything (server, all HD300's, all the networking gear) and things went back to functioning properly. I'll let you know if I problems arise again. The pattern I'm seeing is that everything works great for a few days and then problems start popping up. If there is anything I can do to help you debug this let me know, I'd like to get this rock solid.

Some of my channels have a weak signal which may be a contributing factor.

I've worked a little on some of the things I've noticed that get out of hand when a channel doesn't tune in for a very long amount of time due to no data. The biggest one is limiting the number of times OpenDCT will retry detecting the streams. In your logs, I could tell it was way too aggressive and it lead to eventual saturation of all memory accessible to the JVM instance. This most recent problem was completely centered around one channel from what I could tell.

[tmiranda]

Originally Posted by EnterNoEscape

I've worked a little on some of the things I've noticed that get out of hand when a channel doesn't tune in for a very long amount of time due to no data. The biggest one is limiting the number of times OpenDCT will retry detecting the streams. In your logs, I could tell it was way too aggressive and it lead to eventual saturation of all memory accessible to the JVM instance. This most recent problem was completely centered around one channel from what I could tell.

Had to reboot again today. I can see from the errors there are at least two different channels this time. Log attached.

[EnterNoEscape]

Originally Posted by tmiranda

Had to reboot again today. I can see from the errors there are at least two different channels this time. Log attached.

Your HDHomeRun stopped being accessible from your server. You may very well be having some network related issues. I presume you're running OpenDCT on the same server as SageTV since I can see that SageTV is still talking to OpenDCT, but the HDHomeRun appears to be completely inaccessible for hours. The only other thought I had is maybe the IP address is no longer 192.168.1.25, but even if that were true, OpenDCT would have re-detected it with it's new IP address.

[tmiranda]

Originally Posted by EnterNoEscape

Your HDHomeRun stopped being accessible from your server. You may very well be having some network related issues. I presume you're running OpenDCT on the same server as SageTV since I can see that SageTV is still talking to OpenDCT, but the HDHomeRun appears to be completely inaccessible for hours. The only other thought I had is maybe the IP address is no longer 192.168.1.25, but even if that were true, OpenDCT would have re-detected it with it's new IP address.

Yes, OpenDCT runs of the same server as Sage. How can I tell if my network isn't behaving itself? And why would rebooting the server fix the network issue?

[EnterNoEscape]

Originally Posted by tmiranda

Yes, OpenDCT runs of the same server as Sage. How can I tell if my network isn't behaving itself? And why would rebooting the server fix the network issue?

If you have to reboot the server for everything to start working again that can be an indication that you have a driver problem. If there are newer drivers for your network adapter, I would try upgrading the them.

[tmiranda]

Originally Posted by EnterNoEscape

If you have to reboot the server for everything to start working again that can be an indication that you have a driver problem. If there are newer drivers for your network adapter, I would try upgrading the them.

No newer driver is available.

[EnterNoEscape]

Originally Posted by tmiranda

No newer driver is available.

We've seen things from buggy drivers to bad switches cause problems for people on this thread. Is there any chance you can try another network adapter or plug the HDHomeRun directly into a spare NIC on the server?

[tmiranda]

Originally Posted by EnterNoEscape

We've seen things from buggy drivers to bad switches cause problems for people on this thread. Is there any chance you can try another network adapter or plug the HDHomeRun directly into a spare NIC on the server?

My hardware is old. I could buy a new NIC and see if that helps.

I used the HDHR and Prime a long time with the current hardware without errors. Is OpenDCT just very finicky about the network connection?

[EnterNoEscape]

Originally Posted by tmiranda

My hardware is old. I could buy a new NIC and see if that helps.

I used the HDHR and Prime a long time with the current hardware without errors. Is OpenDCT just very finicky about the network connection?

I don't really have anything better to go on. I don't know if finicky is the right word. I suspect we would be seeing postings similar to yours a lot more often if that was the case. If anything you could say that the network support in Java can be fussy, but that's probably not accurate. I've seen the behavior you're experiencing on other computers from other users, but I have never seen it happen on my own network without the device being unplugged. A fair amount of effort was put into handling all of the usual problems you might run into on a less than perfect network. When OpenDCT says it can't communicate with the device for the 2nd time, it has already taken the entire connection down, ran detection to make sure the IP address has not changed and attempted to reestablish a connection.

I can't recall if I asked this of you already, but have you already tried just restarting the OpenDCT service when you see this issue instead of the entire computer?

Edit: I am working on some pieces of code specific to what you're experiencing and focused on where you're seeing problems.