Spark-Finding which process was killed by Linux OOM killer?

[Spartacus5000]

Hi I test Spark long time, but the problem I described Mgcamd.sh4 which is blocking the tuner Spark, does not solve the problem, eg: the problem of searching for channels when switching around 170 FTA channels and encrypted. Sometimes tuner crashes, and how not to, or searches for channels without a name, or not at all, just off the tuner and on again fixes the problem, but it is best to scan the satellite at the beginning when you turn the tuner Spark, I am very convinced that the reason for suspending the Spark as search without channel names and a total lack of channel search is the "out of memory" or "OOM killer" so it needs a telnet command to display information about what the process was killed by the OOM killer

 

[Mecha]

para conocer mensajes del nucleo, prueba con comando
dmesg
o con el plugin opendebug http://openspa.info/threads/cómo-capturar-el-registro-completo-de-spark.27884/#post-230540

 

[Spartacus5000]

The problem is that I have a laptop but with output to an external monitor, can I connect in such a zone Telnet you can not see? I do not have the RS232 cable and the problem that the laptop has only one output to the monitor :bad I'm sorry, I misunderstood, I'll try this command dmesg, but whether it will show you which processes have killed OOM Killer?

 

[Mecha]

dmesg | grep -i kill

 

[Spartacus5000]

I typed in the telnet command telnet, but nothing appears or something else you need to type? With me it is so by typing:

 

[Mecha]

El comando de filtro | grep -i kill, aplicado sobre la salida del comando dmesg, es para que muestre solo las lineas de la salida del dmesg que contengan la palabra kill, que es la que debe aparecer si algún proceso se ha matado por producirse un error en memoria o para gestionar la falta de memoria por el OOM killer: Out of Memory: Killed process

Tienes que reproducir el problema, y posteriormente ejecutar el comando por telnet, para ver si realmente se ha matado algún proceso por falta o por mala gestión de la memoria del decodificador.

Si no se ha matado ningún proceso en memoria, el problema debe ser un bug de la programación del firmware o de la lista del satélite.

-------
The filter command | grep -i kill, applied to the output of the dmesg command, is to show only the lines of the output of dmesg containing the word kill, which is what should occur if a process has been killed by occur fail to manage memory or forgetfulness by the OOM killer: Out of memory: Killed process

You have to reproduce the problem, and then run the telnet command to see if it really has killed some process by mismanagement or lack of memory of the decoder.

If you did not kill any process in memory, the problem must be a bug of firmware programming or satellite list.

 

[Spartacus5000]

I do not know what the cause is, today I did a test, included plugins: free memory every 5 seconds, oscam + MCAS, after switching around 300 tv channels from time to time it is to stop the system then switches further channels (most visible on Astra 19.2 E German canals), of course, the larger the number of channels, this problem occurs on other satellites, but yes, after approximately 300 channels checked whether the tuner scans transponders and scans OK, so further join channels, probably more than 500 channels, it's been six hours, I checked scan and unfortunately, the tuner scans transponders without any channels my hands fall because he or tuner crashes after about 170-300 channels, but no crashes, but will not scan channels :cry