server freezing issue

[Foggy]

I upgraded from Win 10 1709 to 1803 over the weekend and then to 1909 today.

Once I moved to 1803, the computer started freezing and it appears to happen generally when a scheduled recording starts. No issues when I view a program using Program Guide.

When I say freezing - the screen would be visible but not responsive. The server would no longer respond to pings and the extenders could not connect. The only way to get the server back is to power down/up again.

Nothing in the server event logs.

The sage logs (both sage and sage client) did not say anything either - they just have a time gap between the freeze time and the start of the new log files.

I upgraded OpenDCT to the most recent - no difference. I was at 0.5.28 and moved to 0.5.29
I upgraded SageTV to the most recent - no difference. I was at 9.1.6.something and moved to 9.2.1
I reset the java heap back to 0400H (after the upgrade it was set to 0H)- no difference.

One oddity - by default sage claims it's IP address is 192.168.56.1 - my VirtualBox ip address and not the physical address of the server (192.168.0.10). Disabling the virtualBox ip address so Sage uses the physical address does not seem to help.

Any thoughts?

John

Windows 10, 1909, 64 bit, 24GB AMD Athlon 3 core (circa 2010 build)
SageTV: v9.2.1.826 (32 bit)
OpenDCT: 0.5.29 (32 bit)
Java version: 1.8.0_111
STV file: SageTv7.xml (version 9.2.0.0)
JVM heap size = 236M / 448M / 1038M
HDHomeRun Prime cable tuner
LAV 0.73.1

[Steve20A]

Originally Posted by Foggy

When I say freezing - the screen would be visible but not responsive. The server would no longer respond to pings and the extenders could not connect. The only way to get the server back is to power down/up again.

Similar for me, but my freezes are sporadic and maybe once a month. Disks are spinning, lights on the motherboard are on, but totally unresponsive, even to the keyboard or mouse. TeamViewer is not responsive from other machines. The machine freezes so hard that the reset button does nothing. I, too, have to power down and restart. (I thought the reset button flips a bit in firmware and causes a restart - can firmware freeze?)

Interested to see what experience is out there from others.

[seglertx]

I'm having the same issue but I'm running the 64-bit version of SageTV and it started while I was running Windows 10 Pro version 1903 and has continued with version 1909. The first lockup was October 21st and the last Windows update before they started was on the 11th. Java and NVIDIA updates were installed on the 17th so one of them could be the cause of the problem but I also have another DVR running the 32-bit version of SageTV and the same OS and updates and that hasn't had this problem. I just did a clean installation of Win10 Pro version 1909 and the latest version of SageTV 64-bit on another SSD and while scanning for channels during the setup of SageTV it locked up.

[wnjj]

In addition to the sagetv_*.txt log files that are written from within the application, Java will dump out hs_err_pid* files if the JVM itself crashes (usually due to crashes in native DLL's). Are there any of those from the time/date of the crashes? If so, please post them.

Then again, if it's still running but non-responsive I'm not sure if Java has crashed at all.

[Foggy]

I did make a few changes and my system has not frozen for 30 hours !!

Changes:
OpenDCT - changed sagetv.device.global.required_devices_loaded_count from "0" to "3". (It has been "0" for years.)
Removed the weather plugin from the Installed Plugins / Library section (although it still shows up in the UI Mod and All sections) - probably not the cause.
power cycled my network switch, cable card modem and HD Home Prime boxes.


I do not have any recent java/jvm hs_err files. (for reference, those files are in my C:\Program Files (x86)\SageTV\SageTV folder - your location may differ)

[Foggy]

I also changed the services start type for OpenDCT from Automatic (Delayed Start) to Automatic and left SageTV at Automatic (Delayed Start). This lets OpenDCT get up and running before SageTV needs it.

[Foggy]

3 days without a freeze (good news) but it just froze again (bad news).

There were two extenders playing (one via program guide and the other playing a recording). I happened to be on the server watching a youtube video when it happened. The youtube video stopped, the youtube sound started making a loud racket (probably stutter of some sort?) and a few minutes later the extender playing the recorded video stopped (ah, the power of buffering).

The OpenDCT wrapper.log showed a "PES packet size mismatch" error and the next line in the file had to do with OpenDCT starting up again. However, I see "PES" errors through out the log file without corresponding freezes, so I'm not sure that is the issue.

[KryptoNyte]

It is possible you're experiencing a hardware problem? Hard drive, memory, or network?

[Foggy]

I've wondered about that as well.
So far I've done the following:
- forced a checkdisk on both c:\ and the sagetv media drive (G- both clean.
- tested memory (no issue identified)
- checked all of the drivers to see if any needed to be updated - all were ok.
- I'm not overclocking the cpu or memory

FYI - I had to update the network driver in order to get the Win 10 1803 update to through. I'm wondering if I might purchase a modern e/net card and see if that fixes anything.

I am planning to build another computer late this year or early next year; sooner if this one dies.

The only other thing I've been thinking about is upgrading to the 64bit versions of openDCT and SageTV, but I have to find an open window in the recording schedule first.

[KryptoNyte]

Is it possible to run SageTV in "safe mode with networking" and see if the problem persists?

I had a friend's computer that started blue screening (or whatever Windows 10 does) and it turned out that microsoft had forced a driver update for pcie wireless network device. I pulled the device, and everything was fine. I was disappointed that Windows own repair system couldn't determine that the error was as simple as a driver that it had forced.

I was also surprised at how hard it was to get to Safe Mode on a computer that wouldn't boot.