Reply
Thread Tools
Gusse's Avatar
Posts: 168 | Thanked: 206 times | Joined on Apr 2010 @ Finland
#81
I did comparison between KPv48 and KPv48_Compcache and it seems that RAM and swap usages are both higher with compcache.

What I did: openend microb and did some surfing and monitored Conky. Then left only Conky running and checked RAM and swap after some 5min of idle.

KPv48
RAM: ~170MB
Swap: 90MB

KPv48_Compcache (disksize 98304)
RAM: ~200MB
Swap: 120MB

Wonder why RAM and swap usage is so much higher? I tried to do excatly same things in both cases.

Last edited by Gusse; 2011-08-30 at 11:21.
 
Posts: 1,680 | Thanked: 3,685 times | Joined on Jan 2011
#82
Originally Posted by Gusse View Post
I did comparison between KPv48 and KPv48_Compcache and it seems that RAM and swap usages are both higher with compcache.

What I did: openend microb and did some surfing and monitored Conky. Then left only Conky running and checked RAM and swap after some 5min of idle.

KPv48
RAM: ~170MB
Swap: 90MB

KPv48_Compcache (disksize 98304)
RAM: ~200MB
Swap: 120MB

Wonder why RAM and swap usage is so much higher? I tried to do excatly same things in both cases.
Swap use is higher...Because you now have less RAM...

I imagine the 'swap settings for high I/O' or whatever that thread was called has more appropriate swap settings than say swappolube defaults.

Reasoning is swappolube optimises for minimal swapping (hence low swappiness) and the 'high I/O' settings optimise for lots of little swaps.
__________________
N900: One of God's own prototypes. A high-powered mutant of some kind never even considered for mass production. Too weird to live, and too rare to die.
 

The Following 4 Users Say Thank You to vi_ For This Useful Post:
Posts: 3,074 | Thanked: 12,960 times | Joined on Mar 2010 @ Sofia,Bulgaria
#83
Originally Posted by vi_ View Post
Swap use is higher...Because you now have less RAM...

I imagine the 'swap settings for high I/O' or whatever that thread was called has more appropriate swap settings than say swappolube defaults.

Reasoning is swappolube optimises for minimal swapping (hence low swappiness) and the 'high I/O' settings optimise for lots of little swaps.
Exactly, bellow are my settings

Code:
        echo 30 > /proc/sys/vm/swappiness
        echo 0 > /proc/sys/vm/page-cluster
        echo 512 > /sys/block/mmcblk0/queue/nr_requests
        echo 512 > /sys/block/mmcblk1/queue/nr_requests


        echo 0 > /sys/block/mmcblk0/queue/iosched/slice_idle
        echo 0 > /sys/block/mmcblk1/queue/iosched/slice_idle
        echo 32 > /sys/block/mmcblk0/queue/iosched/quantum
        echo 32 > /sys/block/mmcblk1/queue/iosched/quantum
        echo 80 > /sys/block/mmcblk0/queue/iosched/fifo_expire_sync
        echo 80 > /sys/block/mmcblk1/queue/iosched/fifo_expire_sync
        echo 180 > /sys/block/mmcblk0/queue/iosched/fifo_expire_async
        echo 180 > /sys/block/mmcblk1/queue/iosched/fifo_expire_async
... and swap on microsd
 

The Following 8 Users Say Thank You to freemangordon For This Useful Post:
Gusse's Avatar
Posts: 168 | Thanked: 206 times | Joined on Apr 2010 @ Finland
#84
Originally Posted by vi_ View Post
Swap use is higher...Because you now have less RAM...
Of cource, what was I thinking

If we run out of RAM, could this cause lack of responsiveness due to heavy swapping activity? Just trying to understand what is causing quite big loading conditions in my system when using Compcache.
 
F2thaK's Avatar
Posts: 4,365 | Thanked: 2,467 times | Joined on Jan 2010 @ Australia Mate
#85
I did the following as root in xterm and got errors

Code:
insmod /lib/modules/current/ramzswap.ko disksize_kb=120512
swapon /dev/ramzswap0
swapoff /dev/mmcblk0p3; swapon /dev/mmcblk0p3
insmod /lib/modules/current/ramzswap.ko disksize_kb=120512 = -1 invalid module format
swapon /dev/ramzswap0 = no such file/dir
 
Posts: 3,074 | Thanked: 12,960 times | Joined on Mar 2010 @ Sofia,Bulgaria
#86
Originally Posted by F2thaK View Post
I did the following as root in xterm and got errors

Code:
insmod /lib/modules/current/ramzswap.ko disksize_kb=120512
swapon /dev/ramzswap0
swapoff /dev/mmcblk0p3; swapon /dev/mmcblk0p3
insmod /lib/modules/current/ramzswap.ko disksize_kb=120512 = -1 invalid module format
swapon /dev/ramzswap0 = no such file/dir
use the kernel from the first post, ok?
 
TMavica's Avatar
Posts: 2,021 | Thanked: 1,060 times | Joined on Apr 2010 @ Hong Kong
#87
when swap full, phone become laggy
 
Posts: 3,074 | Thanked: 12,960 times | Joined on Mar 2010 @ Sofia,Bulgaria
#88
what disksize_kb are you using. Also do you use some speedpatches?
 
TMavica's Avatar
Posts: 2,021 | Thanked: 1,060 times | Joined on Apr 2010 @ Hong Kong
#89
Originally Posted by freemangordon View Post
what disksize_kb are you using. Also do you use some speedpatches?
use yours recommend size 120512, speedpatch? I removed
 
Posts: 3,074 | Thanked: 12,960 times | Joined on Mar 2010 @ Sofia,Bulgaria
#90
And you are sure speedpatch uninstalls correctly? No remnants? Anyway we are still trying to discover what are the best settings to be used with compcache.
For now you may try (in xterm as root)

echo 20 > /proc/sys/vm/swappiness
echo 2 > /proc/sys/vm/page-cluster

hoping that other vm settings are stock. if that not help then I can bet you have some remnants from some optimization script that have messed your /proc/sys/vm/ (and who knows what else) values. In case lagging continues would you please do cat /proc/sys/vm/* and copy the result here
 

The Following User Says Thank You to freemangordon For This Useful Post:
Reply


 
Forum Jump


All times are GMT. The time now is 12:56.