22 replies to this topic

[HWTest]

I've encountered a crash (without a crash log) on my new Ultimo 4k running OpenPLi 6.2.

I was watching a recording, one recording was already running a three were starting. At this moment the playback stopped and I got a no free tuner warning, then the spinner and a reboot (probably just enigma restart). After the reboot, again no free tuner warning but all 4 recording successfully restarted. Sadly after this, Oscam was in some strange state and the new recordings were scrambled.

Is this "normal" (as expected) and I should check that recording don't start at the same time?

 

PS: I remember on VTI there was a choice "don't allow recordings start at the same time" or something like that. I've never set it to on and never had problems with multiple recordings starting at the same time - but that was on the Duo2 with regular (non FBC) tuners. Maybe it's a FBC tuners specific problem.

[WanWizard]

I know it has been worked on for OpenPLi 7, but I can't remember if it was completely fixed or not.

[ccs]

The issue was discussed in this very long thread....

 

https://forums.openp...ous-recordings/

[littlesat]

In openpli 7.0 we have a work-a-round included that solved this topic... (when someone did not revert it)... it does solve the diseqc switch issue and also exact this one...

Edited by littlesat, 7 January 2019 - 16:32.

[Rob van der Does]

That 'long thread' was not about a crash, but about failing (zero length) recordings.

[littlesat]

I speculate this is what here is meant with crash without crashlog

[ccs]

That 'long thread' was not about a crash, but about failing (zero length) recordings.

But a workaround for timers starting at the same time was discussed.

[HWTest]

I don't know about the "first" recordings (before the crash), because the box crashed immediately after they started, so I deleted them without checking.

The "second" recordings (after the crash) had normal sizes but were scrambled because Oscam didn't survive the crash and I wasn't able to offline decode them afterwards (a day later).

 

Starting to read the very long thread ...

[Rob van der Does]

That 'long thread' was not about a crash, but about failing (zero length) recordings.

 
 

I speculate this is what here is meant with crash without crashlog

I don't think so. OP mentions 'no free tuner', which was absolutely no issue in the mentioned 'long thread & work-around'.

[anudanan]

I often have different timers with the same start time most at prime time and that works on 7.0rc

Sometimes in the past I have seen a driver crash by starting timers with the same start time on 6.1

Box is uno4kse

I‘ve no driver crashes seen on 7.0 since today and my wife make very much recordings

[Erik Slagter]

The issue discussed in the other thread was for a very specific case, with multiple FBC tuners having SCR tuners assigned. I don't think that is the case here.

 

The next time this issue plays, I suggest you login, type "dmesg" and at the bottom you'll see something like "kernel OOPS" or "null pointer reference" and then a stack trace.

 

Which means the drivers have crashed. Only solution is to reboot.

[ccs]

I only mentioned the other thread because of this comment....

 

 

Are you sure this works... I also have issues with timers starting on the same time (without jess/unicable).... and this ‘work a round’ looks really simple

 

[HWTest]

Hmmm, so what should I do?

 

Check that no timers start at the same time and take it as normal?

Wait for OpenPLi 7?

Try to reproduce the problem and post the dmesg output?

[anudanan]

From m point of view take openpli7.0rc. There are new driver included and for me the driver are better with fixed crashed. With my uno4kse I have a same FBC tuner as you in your ultimo4k

[ColinB]

Hmmm, so what should I do?

 

Check that no timers start at the same time and take it as normal?

Wait for OpenPLi 7?

Try to reproduce the problem and post the dmesg output?

 

Move to OpenPLi 7.0-rc

 

No point in waisting time on old image.

[anudanan]

Today my uno4kse start four timer at 20:04 and four timer at 20:05 with no crashes.
I think 7.0rc is more stable than 6.x on VU+

[anudanan]

I don‘t know if they started with seconds delay. But maybe

[anudanan]

I‘ve startet 3 more timer at the same timer and saw in the timerlist they start with a second delay timer1, delay , timer2, delay , timer3

So you are right

[ccs]

I deleted my last post because I could only see timers at the hour and minute level, no detail of seconds that I could find.

[HWTest]

OK, is the 7.0 RC stable enough for daily use?