Posts by majoly

My usb WIFI dongle was hooked up to my powered usb hub.

I connected the dongle to a usb3 port instead.

Problem did not reappear.

Thanks!

Ok my setup is a Server running Libreelec on a Pi 4 4gb. Its running tvheadend latest version. 9.1.901.444

Once in a while I am watching live tv and the image if frozen for between 5-30 seconds and after it starts playing again.

This happens mostly when watching live TV.

I installed the necessary addons to start a basic diagnostic to understand whats going on when its frozen.

When running normally, I see in my Network I/O using NMON that its transmitting fine and all my CPUS are mostly IDLE.

When it freezes, Half my CPU's are on WAIT and the Transmitted info is at 0kb/s.

I checked the logs and nothing was printed at that time.

No recording were active at that time and Timeshift is active.

Any idea on what to check next?

Thanks

I may be having the same issue.

My server is now a Pi 4 and when I watch live tv with my other Pi in my living room, sometimes it stops playing for a few seconds, audio is out of sync sometimes and sometimes it gets pixelated.

I though it could be a network issue but with my 3B+ I had previously, it wasent an issue.

Found the solution.

Used the guide.xml generated by my WebGrabPlus service instead and have set the External Epg Grabber instead.

I then had to add a cronjob that does a curl to my xmltv.sock.

Now my tv schedule is working!

Is there a knows issue with latest version where if the signal quality is weak, image could freeze and never come back on its own and you have to switch channel to make it come back?

With TvHeadEnd 4.2.8 I did not get that issue.

I came back to my 3B+ , with latest version 4.3.1802, same result

Quote from Mario77

So probably guide.xml should be in the same folder of WebGrab++.config.xml

Are you using TvHeadend the normal version from Libreelec repo or the docker version?

In the normal version you just point to the guide.xml from the tvheadend addon settings and you enable the Internal Xmltv Grabber from Tvheadened UI.

The docker version I never used sorry so not sure if it's the same or you might need to make some script/cronjob to make it work..

Docker version.

Normal version of TvHeadend I had no problem configuring this.

For the Docker version I have no clue where to set that...

There is a <filename>guide.xml</filename> by default meaning that wherever the executable is, the xml will be written there.

I got my guide.xml, where im lost is where to set in TvHeadEnd where to get that guide.xml

I installed TvHEadEnd 4.3 and WebGrab+ from linuxserver.io repo.

In multiple threads I saw, I was never able to determine where to put the guide.xml or specify its location to TvHeadEnd...

Any ideas?

Opted for 4B 4GB

Thanks!

Ok my issue never came back.

I feel dumb that I always presumed I was running latest version of TVHeadEnd which wasn't the case.

I have other issues but not related to this one.

Thanks for the help.

My Server Pi4 4gb

My Tv with this issue is running on a Pi 3B

When TV is live, it sometimes happens that the image becomes pixelated and freezes but sound is still working.

My server was previously a 3B+ and I had bunch of other issues but not this one.

Since im running on the same Pi 3B as before to watch TV, is there a setting that I need to change on the Pi 4B+ for that?

Thanks

Quote from Iridium

Can you run a program to output free -m every hour or so and when a crash occurs, post the output.

Also what other addons, programs are you running outside of normal LE?

No other program.

This server is mostly dedicated for tvheadend. Its not even connected to a TV so there isnt any other addons installed.

I realized something tho that I don't know why I had never noticed before but TVHeadend version I had installed was the one in libreelec repo which was 4.2.somethin.

I saw that the docker linuxserver.io one was 4.3 so I decided to try it out.

The problem I had did not appear since then.

I will not say for now the problem is solved since it sometimes can be difficult to reproduce.

Week-end is coming so ill have some time to make it crash again if I can!

Quote from smp

Try to add this to /storage/.kodi/addons/service.tvheadend42/bin/tvheadend42.start, below the . /etc/profile line:

Code

export MALLOC_ARENA_MAX=1
export MALLOC_MMAP_THRESHOLD_=131072

Tried this modification and no difference. Same problem occurs.

I will re-try the swapfile modification

In a previous thread I started asking for an alternative for a 3b+, you made this same suggestion and tried it out.

With the 3b+ it did not work and it became more unstable for some reason.

I will give it a shot with my Pi 4 also we never know.

Little update.

Since everything I was asked to try failed, i decided to try a Raspberry Pi 4 4GB version in hope there really is a memory issue and that it will solve it.

Sadly the issue is still present even with 4gb of ram.

Before the problem occured, the memory was steady high at 3.5gig of ram free in average and when the problem started occuring, I was able to see the memory going down rapidly to 90mbit until it crashed again. It went from 3.5gig to 90m in about a minute.

From this point I guess it really its a memory leak going on with tvheadend.

I have found similar post about this but could not get any resolution.

Well I got tired of the problems and disabled the swapfile.

I was starting to get networking issues and instability that I have not seen before.

No new crash log giving me info on what is happening with swapfile active.

Now that it is disabled, i'm back to my old instability with killing tvheadend, out of memory.

Let's say that I want to try the Odroid N2, which version do I install on it? The same as the Odroid C2 on the download page?

Still having issues.

Hard to say if same thing is still happening.

All I know is that new problems started occuring since i enabled swapfile...

Requires sometimes to reboot my Pi multiple times to be visible on network.

Maybe a coincidence I do not know.

Once ill have a decent crash log ill post it somewhere else from here

Just enabled it and now it's the waiting game.

It only happens like once a day around the same time.

If after a week I do not see it again crash, guess problem will be finally solved