Force File Swap reduce RAM

Agreed, but what can we or the developers do to get this under control ? ever increasing ram use is going to cause a crash...
 
Agreed, but what can we or the developers do to get this under control ? ever increasing ram use is going to cause a crash...
I don't know how to solve.
Maybe you can schedule a reboot of your box some minutes before epg refresh, so your epg will always be light and new.
 
And that to say four recordings without freezees ?

With my solution to force RAM ; Two (2) recording SAT and Two (2) DTT ?


Best regards .
CalamaresRellenos .
 
Maybe you can schedule a reboot of your box some minutes before epg refresh, so your epg will always be light and new.
Tamaki is a wise man :)
and
if further clean is needed is enough crond (or an "invasive" free ram continous verify then give the) appropriate comand ....
i'd like to use it just after a full backup
 
I will conclude here now that adding a swapfile will help prevent EPG crashes but only until swapspace runs out, if that happens it will crash! using values that force swapfile use will only make that crash happen sooner, on the other hand a swapfile makes the box work slower day by day.

So whats next?

1. Enigma2 needs to be changed to deal with EPG space use, either by clearing epg space before each load or by moving epg data from ram to some other device with an option to 'delete' it via a script.

2. I've played around with various ways but the only way to keep the box running is to reboot E2, however there is no default way to reboot once or twice a week because the crontab daemon does not handle weekdays properly, so here is my workaround for a boot once or twice a week:

Use ftp or samba and go to \etc\bhcron
Here you will find 2 files: 'bh.cron' and 'root' which both contain the same values (bug?)
In both files put the lines:
Code:
20 5 * * * if [ `/bin/date +%a` = "Wed" ]; then /usr/bin/Blackholecmd restartenigma2; fi
20 5 * * * if [ `/bin/date +%a` = "Sun" ]; then /usr/bin/Blackholecmd restartenigma2; fi
22 5 * * * /usr/bin/Blackholecmd standby
This will reboot E2 twice a week on a Wednesday and on Sunday at 05:20, 2 minutes later it will go to standby mode.
Use BH Blue,Blue,BHCrontab to check if BHcron see's this.
(Note: the 5th star, dayofweek value, does not work on the VU+ which explains the if-fi solution)
For more ways to use the +%a values see:
Code:
http://www.go2linux.org/date-linux-command-line-set-the-linux-time
About 5 minutes later after standby you can schedule CrossEPG to run.
 
I will conclude here now that adding a swapfile will help prevent EPG crashes but only until swapspace runs out, if that happens it will crash! using values that force swapfile use will only make that crash happen sooner, on the other hand a swapfile makes the box work slower day by day.

So whats next?

1. Enigma2 needs to be changed to deal with EPG space use, either by clearing epg space before each load or by moving epg data from ram to some other device with an option to 'delete' it via a script.

2. I've played around with various ways but the only way to keep the box running is to reboot E2, however there is no default way to reboot once or twice a week because the crontab daemon does not handle weekdays properly, so here is my workaround for a boot once or twice a week:

Use ftp or samba and go to \etc\bhcron
Here you will find 2 files: 'bh.cron' and 'root' which both contain the same values (bug?)
In both files put the lines:
Code:
20 5 * * * if [ `/bin/date +%a` = "Wed" ]; then /usr/bin/Blackholecmd restartenigma2; fi
20 5 * * * if [ `/bin/date +%a` = "Sun" ]; then /usr/bin/Blackholecmd restartenigma2; fi
22 5 * * * /usr/bin/Blackholecmd standby
This will reboot E2 twice a week on a Wednesday and on Sunday at 05:20, 2 minutes later it will go to standby mode.
Use BH Blue,Blue,BHCrontab to check if BHcron see's this.
(Note: the 5th star, dayofweek value, does not work on the VU+ which explains the if-fi solution)
For more ways to use the +%a values see:
Code:
http://www.go2linux.org/date-linux-command-line-set-the-linux-time
About 5 minutes later after standby you can schedule CrossEPG to run.


This is a very good solution. Thanks.
 
OK, so do we now have advise for general usage? It would be great if some one could summarize the best bits in one post here. :)
 
Back
Top