You are not logged in.
[root@li83-30 ~]# uname -a
Linux li83-30 2.6.32-x86_64-linode11 #1 SMP Sat Dec 5 16:55:26 UTC 2009 x86_64 Intel(R) Xeon(R) CPU
L5520 @ 2.27GHz GenuineIntel GNU/Linux
[root@li83-30 ~]# free -m
total used free shared buffers cached
Mem: 343 90 252 0 8 48
-/+ buffers/cache: 33 309
Swap: 511 0 511
[root@li83-30 ~]# pacman -Syy
:: Synchronizing package databases...
core 36.1K 249.7K/s 00:00:00
extra 441.8K 64.6K/s 00:00:07
community 370.8K 66.7K/s 00:00:06
[root@li83-30 ~]# free -m
total used free shared buffers cached
Mem: 343 180 162 0 46 82
-/+ buffers/cache: 51 292
Swap: 511 0 511
[root@li83-30 ~]#
Offline
No, it's because 64bit applications utilize more ram than the 32bit ones.
Offline
I loaded my linode vps with 64-bit Debian 5.0 and tried comparing memory usage of apt-get with pacman. 'apt-get update' used around 50 MB RAM while 'pacman -Syy' used around 90 MB RAM.
li83-30:~# uname -a
Linux li83-30 2.6.18.8-x86_64-linode10 #1 SMP Tue Nov 10 16:29:17 UTC 2009 x86_64 GNU/Linux
li83-30:~# free -m
total used free shared buffers cached
Mem: 360 39 321 0 1 10
-/+ buffers/cache: 27 333
Swap: 511 0 511
li83-30:~# apt-get update
Get:1 http://security.debian.org stable/updates Release.gpg [835B]
Get:2 http://security.debian.org stable/updates Release [40.8kB]
Get:3 http://security.debian.org stable/updates/main Packages [338kB]
Get:4 http://security.debian.org stable/updates/contrib Packages [20B]
Get:5 http://mirror.cc.columbia.edu lenny Release.gpg [1032B]
Get:6 http://mirror.cc.columbia.edu lenny Release [73.6kB]
Ign http://mirror.cc.columbia.edu lenny/main Packages/DiffIndex
Get:7 http://mirror.cc.columbia.edu lenny/contrib Packages [86.8kB]
Get:8 http://mirror.cc.columbia.edu lenny/main Packages [6848kB]
Fetched 7389kB in 10s (738kB/s)
Reading package lists... Done
li83-30:~# free -m
total used free shared buffers cached
Mem: 360 87 272 0 1 57
-/+ buffers/cache: 28 331
Swap: 511 0 511
li83-30:~#
Offline
Using free before and after a program ran is an incredibly naive way of measuring memory usage... so much, that the numbers you posted are quite meaningless.
Offline
I apologize for my ignorance. I am trying to get a very rough picture here. Can you explain why -/+ buffers/cache: shows a big difference before and after running "pacman -Syy".
Why does "apt-get update" does not show that kind of difference in -/+ buffers/cache: before and after running it?
I thought -/+ buffers/cache: is supposed to show the amount of really used RAM after subtracting the sizes of buffers and caches from the total memory used.
Offline
Why does it matter?
You buy RAM to use it.
Offline
I think a reason might be, that "apt-get update" is more comparable with "pacman -Sy", rather than "pacman -Syy"
Offline
I am not familiar with internals, but it may be a time/memory trade off somewhere in pacman, so that it executes faster.
[git] | [AURpkgs] | [arch-games]
Offline
I apologize for my ignorance. I am trying to get a very rough picture here. Can you explain why -/+ buffers/cache: shows a big difference before and after running "pacman -Syy".
Why does "apt-get update" does not show that kind of difference in -/+ buffers/cache: before and after running it?
I thought -/+ buffers/cache: is supposed to show the amount of really used RAM after subtracting the sizes of buffers and caches from the total memory used.
Start htop before you run pacman, and look at the ram usage of all the processes running. Run pacman and see what changes.
Last edited by Mr.Elendig (2010-01-28 13:42:38)
Evil #archlinux@libera.chat channel op and general support dude.
. files on github, Screenshots, Random pics and the rest
Offline
I apologize for my ignorance. I am trying to get a very rough picture here. Can you explain why -/+ buffers/cache: shows a big difference before and after running "pacman -Syy".
Why does "apt-get update" does not show that kind of difference in -/+ buffers/cache: before and after running it?
I thought -/+ buffers/cache: is supposed to show the amount of really used RAM after subtracting the sizes of buffers and caches from the total memory used.
After running a process, all the memory it uses is freed. I.e. not in the "used" column... Some stuff might sit around in the cache to speed up the next run but that is freed when needed too.
But still, the cache usage tells up absolutely nothing. For the same cache usage, a program might have read a big chunk of data or it might have read and free lots of small chunks of data. One will use substantially more memory than the other. So, cache usage is meaningless.
A google search will show you how to actually measure the maximum memory usage of a program.
Offline
IIRC, Dan (toofishes) has done a few pacman revisions over the years to seriously whittle down its memory usage, which by now should be quite relatively low I would say.
Offline