how can i reduce RAM usage on my blackhole image

Matrix10

Matrix10

Administrator
I had never crash of the image, unless there is a bug
in plugins, skin or whatever , and apart from a lot of EPG providers.
In normal use and maximum of +- 3 EPG providers I have never image crash
and without to use a swap file which I never used only for test
With test that I have make with SWAP
I was not convinced that it can prevent crash of the image with to many
EPG providers.
Why ?
In my opinion, because enigma does not allow overloading of memory with the EPG and as protection causes crash enigma and cleaning the memory.
I have test image (I will not mention the name) that prevents enigma crash with many EPG
Pumping epg to a maximum in memory and it causes another problem which is that image
not working properly (Or does not respond to remote commands ) , and that is a bigger problem than enigma crash .
 
B

bh4me2

Vu+ Newbie
Yes Matrix I know, you have said that a few times, however I have had many crashes with only 2 epg providers en 3 plugins, nothing mayor, no big epg load and still it crashed. Now I have the lsk plugin and a small usb-swapfile and so far no crashes, full 7 day epg gets loaded everyday, I can get statistics (blue-orange), record, timeshift, etc... and no crash.

It is however strange that the moment my swapfile got used is about the moment it usually crashed before.

If there is a way to see whats in the swapfile maybe we can figure out whats running out of ram and then figure out why.
 
Matrix10

Matrix10

Administrator
I really do not know ,
because I do not download EPG every day , but every few days.
So I do not have these problems.

I have tested previously SWAP with a large number of EPG providers
and swap failed to prevent crash.So I do not use swap because I do not need.
 
a911

a911

The BH Lover
bh4me2 said:
Well I've added lsk 03 and at first it seems it's not doing anything but after a few hours I get this:
Mem: 88192K used, 53288K free, 0K shrd, 432K buff, 13412K cached
CPU: 0.8% usr 0.4% sys 0.0% nic 97.0% idle 0.0% io 0.0% irq 1.5% sirq
Load average: 0.79 0.22 0.07 2/82 5344
So the Cache does remain used but gets reduced, swap file still active but still not used.

View attachment 4103
Click to expand...

USB in use: 100%.......:chair:!

What do U have there......:shocked:?
Get bigger one (USB), may be your problem will go away....;)!

GdLuK
a911
 
B

bh4me2

Vu+ Newbie
a911: Usb in use 100% because the swapfile is located on this usb so this is correct.
However last night I had a BIG crash, the whole box hung, had to flip the power switch, after restart epg is empty, looked for a crashlog but it was empty (zero bytes), crossepg did run, the clock on the hanging vu+ was showing 1 minute after crossepg was finished, crossepg database was processed ok so it must have crashed at loading epg data into E2 which is the last step of crossepg. Swapfile is empty again.
Matrix: how do you schedule epg update once or twice a week ? or do you run crossepg manually ?

Would it help if we added something that clears epg space in E2 before crossepg loads new epg ? if so how would I go about that ?
 
angelofsky1980

angelofsky1980

BlackHole Driver Specialist
bh4me2 said:
a911: Usb in use 100% because the swapfile is located on this usb so this is correct.
However last night I had a BIG crash, the whole box hung, had to flip the power switch, after restart epg is empty, looked for a crashlog but it was empty (zero bytes), crossepg did run, the clock on the hanging vu+ was showing 1 minute after crossepg was finished, crossepg database was processed ok so it must have crashed at loading epg data into E2 which is the last step of crossepg. Swapfile is empty again.
Matrix: how do you schedule epg update once or twice a week ? or do you run crossepg manually ?

Would it help if we added something that clears epg space in E2 before crossepg loads new epg ? if so how would I go about that ?
Click to expand...

USB 100% in System Info mean that the device is FULL and doesen't have free space on it!!!!
Please empty the USB device or change it with a bigger one.
 
angelofsky1980

angelofsky1980

BlackHole Driver Specialist
bh4me2 said:
Please read:
http://www.vuplus-community.net/board/threads/swap-file-size-psition.2169/

View attachment 4129

Creating a swapfile on a USB device will list this device as fully used if the size of the swapfile is as big as the USB device. The 'free' command lists how much swap is in use (0%) and how much is available (233472) which is the same as the size of this USB device, simply because the entire USB is set as swapfile. You do understand linux or not ?
Click to expand...

I understand Linux very well.

Swapfile is a kind of memory who is summed to RAM memory.
if you use ALL of your USB stick as SWAP, the device cannot be used for any other kind of purposes like MultiBoot, Picons and EPG.
Reduce your swapfile size and you should solve your problem of CrossEPG crashes.
 
Matrix10

Matrix10

Administrator
Matrix: how do you schedule epg update once or twice a week ? or do you run crossepg manually ?
Click to expand...

Problem with CrossEPG is that you can turn on only>> once a day
As far as I know,or you must use manual.
We can talk for days about the EPG and the crash of enigma
I explained why there is a crash enigma
and I think that we have reserved memory space for the EPG
pumping EPG greater than that caused instability or image collapse.
I do not know how this memory is calculated for the EPG use .
The problem is also that the error in the EPG data can also cause crash enigma
not only a problem of memory space.
As far as I know no one has yet found a proper solution to this problem.
And all the other experiments with SWAP did not yield satisfactory results.
Not in my testing.

SWAP may help the work of other memory
users' but not EPG, although it seems that way
It is my opinion.

If I create the swap at the same time download more EPG providers
I will again have crash of enigma despite the size of a swap file
 
Dzoni

Dzoni

BH Lover
Just LSK free memory can get minimal help because remove unused cash and can give to us possibiluty for 1 more EPG provider. That is a maximum.
 
tobler

tobler

Vu+ User
the thread starter left this discussion, maybe is better to start another one

you (all) talking about another thing
 
a911

a911

The BH Lover
bh4me2 said:
a911: Usb in use 100% because the swapfile is located on this usb so this is correct.
However last night I had a BIG crash, the whole box hung, had to flip the power switch, after restart epg is empty, looked for a crashlog but it was empty (zero bytes), crossepg did run, the clock on the hanging vu+ was showing 1 minute after crossepg was finished, crossepg database was processed ok so it must have crashed at loading epg data into E2 which is the last step of crossepg. Swapfile is empty again.
Matrix: how do you schedule epg update once or twice a week ? or do you run crossepg manually ?

Would it help if we added something that clears epg space in E2 before crossepg loads new epg ? if so how would I go about that ?
Click to expand...

What size of USB you have?
Plus 128 MB ,of extra memory for swab file, is more than enough!
 
B

bh4me2

Vu+ Newbie
Lincsat said:
You could daily cron a re-boot before the CrossEPG, that would clear the EPG data in RAM
Click to expand...
Yes if cron actually worked, I've tried cron but cron isn't working for some reason. But on another note a full reboot does not clear epg, only a full power off clears it, I've done quite a few full reboots this weekend and my epg is still there.
 
B

bh4me2

Vu+ Newbie
a911 said:
What size of USB you have?
Plus 128 of extra memory for swab file is more than enough!
Click to expand...
256mb stick used fully as swap space, but I'm going to get a faster one today as this one was one I was given (no name gift crap).
 
a911

a911

The BH Lover
bh4me2 said:
256mb stick used fully as swap space, but I'm going to get a faster one today as this one was one I was given (no name gift crap).
Click to expand...

The USB always reserve few MB for its drives,,I think....!
So go for 2 giga ,its more than enough,,,& lets know the results!
 
mirkoslav

mirkoslav

Vu+ Newbie
Hi all, I had same issue with Vu+ Duo, because I download 8 providers at the moment through CrossEPG. Created 32Mb Swap file on HDD and defined scenario in Cron that runs every day early in the morning - 1. Run script to delete all files in CrossEPG folder, 2. Restart box, 3. (in time frame here is schedule for CrossEPG download), 4. Restart box, 5. Go to standby
This way things look great, box is fast, memory remains around 90% and swap file seems unused (any time I looked it was 0%)

But I have a question here, related to swap. My flash memory is around 45% in use when I load everything needed. Question to team is, would it be possible to allow swap creation in flash memory? Half of it is empty and we need to create swap on (slower) hard disc or USB. As said, it can be a small size swap, but it would allow extending RAM through much faster internal flash memory sitting unused.

I will volunteer testing if needed :)
 
B

bh4me2

Vu+ Newbie
As far as I can tell it is not the size of the epg file, crashes happen with just about any size. Expanding flash is a bad thing as flash gets written to way too often (flash has a limited lifetime with writing). What we need is to fully separate epg from e2 processes on hdd/usb or swapspace and just keep pointers in ram as to where epg resides, this will also solve the loading-epg issue as it does not need to be loaded anymore into ram/flash avoiding all crashes.

Ea. its there, this is where it starts and this is where it ends, whatever is between start and end is data only, add some data consistency checks (format of epg data check) and there you will have it, a non-crashable epg system.
 
You must log in or register to reply here.
Top