anticapitalista
Posts: 5,955
Site Admin
Joined: 11 Sep 2007
#91
I just tested this using user-management to set up new user (demo psswd=demo)
Logout - login to demo
Logout - login usual user used user-management to remove demo.
No issues no errors.
Posts: 1,062
Dave
Joined: 20 Jan 2010
#92
hmm, not sure what to say.
My m12 core build did not have the error, but running antix full prefinal in vbox (fresh install) had the issue no matter what the username is.

i do not know what rc was.running, maybe he can post later.

and just in case it has something to do with the new user management script (even though the old had pthe problem for me) here is the download folder

========= SCRAPER REMOVED AN EMBEDDED LINK HERE ===========
url was:"http://daveserver.info/antiX/development"
linktext was:"http://daveserver.info/antiX/development"
====================================

oh i was testing using icewm rox as my session as well
Posts: 765
rust collector
Joined: 27 Dec 2011
#93
I was using M12 full icewm rox.... Oh, and wicd has been removed
Posts: 765
rust collector
Joined: 27 Dec 2011
#94
New try today, the script works fine in fluxbox...no errors...
Posts: 765
rust collector
Joined: 27 Dec 2011
#95
And with rox-fluxbox... the same error pops up...
So, it is probably something to do with rox?
anticapitalista
Posts: 5,955
anticapitalista
Site Admin
Joined: 11 Sep 2007
#96
Ok, Dave I see the issue now.

/usr/bin/dbus-daemon --fork --print-pid 5 --print-address 7 --session

doesn't get killed on logout, so if user logins in and out regularly, we get multiple pid for

/usr/bin/dbus-daemon --fork --print-pid 5 --print-address 7 --session

Now if we now a way to kill

/usr/bin/dbus-daemon --fork --print-pid 5 --print-address 7 --session
we can add it to the first line of usr/local/bin/logouthelper.sh

Code: Select all

killall /usr/bin/dbus-daemon --fork --print-pid 5 --print-address 7 --session 
doesn't do it.
Posts: 1,062
Dave
Joined: 20 Jan 2010
#97
placing
USER=`whoami`;
before you start closing any sessions, and then
sleep 2 && pkill -u"$USER"
after closing all sessions may do it.

however if the logout script is accidently run as root it would kill your whole system

Maybe you can solve that by doing a root check,
ROOT="$UID"
if ["$ROOT" !="0" ]; then
sleep 2 && pkill -u"$USER"
fi

There are more options to, here is a link

========= SCRAPER REMOVED AN EMBEDDED LINK HERE ===========
url was:"http://www.unix.com/shell-programming-scripting/117204-kill-all-processes-belonging-one-user.html"
linktext was:"http://www.unix.com/shell-programming-s ... -user.html"
====================================

find the post by zxmaus (about 3/4 down the page second last post)
Posts: 162
drg
Joined: 22 Feb 2010
#98
antiX-486-base Pre-Final Edelweiss testing
K6/266MHz/128Mram. with kernel 3.4.5-antix.1
swap partition 400M (gparted), inxi sees as 0.42GB, conky as 399M without zram or 508M with zram (adds the swap -1 -2 -3).
On boot up zram did “Creating two devices” ie 2 swapspaces each 55648 KiB, swap -2 and -3.

For testing needed some ram grabbing gui apps, so I installed abiword, gnumeric, yelp, spacefm, pysolfc, osmo, synaptic, LOwriter, LOcalc.
Am going to keep loading more applications on top of each other. Using inxi -F and conky for measuring ram/swap usage. Note that with idling or any event happening, the figures change a bit, so they are just a snapshot of that moment.


****With rox-fluxbox and roxterm:
$ inxi -F (hereafter edited for the good parts only)
System: Host: antiX1 Kernel: 3.4.5-antix.1-486-smp i586 (32 bit)
Desktop: Fluxbox 1.3.2 Distro: antiX-base Edelweißpiraten 26 May 2012
Machine: System: IBM product: 2153E2N version: 0000000000000000
Mobo: 0000000000000000 model: 0000000000000000 version: 0000000000000000
Bios: ACER version: V3.2 V70EN1X date: 08/18/98
CPU: Single core AMD-K6 3D (-UP-) cache: 64 KB flags: (-) clocked at 266.033 MHz
Graphics: Card: Advanced Micro Devices [AMD] nee ATI 3D Rage Pro AGP 1X
X.Org: 1.12.1.902 drivers: ati,mach64 (unloaded: fbdev,vesa) Resolution: 832x624@75.0hz
GLX Renderer: Gallium 0.4 on softpipe GLX Version: 3.0 Mesa 8.0.3
Audio: Card: Ensoniq 5880B [AudioPCI] driver: snd_ens1371 Sound: ALSA ver: 1.0.25
Drives: HDD Total Size: 6.5GB (-) 1: id: /dev/sda model: FUJITSU_MPE3064A size: 6.5GB
Partition: ID: / size: 5.6G used: 2.5G (46%) fs: ext3 ID: swap-1 size: 0.42GB used: 0.00GB (0%) fs: swap
ID: swap-2 size: 0.06GB used: 0.00GB (0%) fs: swap ID: swap-3 size: 0.06GB used: 0.00GB (0%) fs: swap
Sensors: None detected - is lm-sensors installed and configured?
Info: Processes: 64 Uptime: 4 min Memory: 45.9/120.8MB Client: Shell inxi: 1.8.4

****Adding abiword:
ID: swap-1 size: 0.42GB used: 0.00GB (0%)
ID: swap-2 size: 0.06GB used: 0.00GB (3%)
ID: swap-3 size: 0.06GB used: 0.00GB (3%)
Memory: 57.4/120.8MB
--conky ram 54.2M/121M -44%, swap 3.04M/508M -0%

====> note swap -2/-3 kicked in already at 3%
at idle settled to ram 59.4M -49%, swap 3.06M -0%
after more idle got ram 58.5 -48%, swap 4.78 -0%

****Adding Control Center/ keys invoking geany (just to see the fluxbox hotkeys):
ID: swap-1 size: 0.42GB used: 0.00GB (0%)
ID: swap-2 size: 0.06GB used: 0.00GB (6%)
ID: swap-3 size: 0.06GB used: 0.00GB (6%)
Memory: 67.7/120.8MB
--conky ram 64.2M -54%, swap 6.21M -1%

****Adding spacefm:
ID: swap-1 size: 0.42GB used: 0.00GB (0%)
ID: swap-2 size: 0.06GB used: 0.01GB (13%)
ID: swap-3 size: 0.06GB used: 0.01GB (13%)
Memory: 78.5/120.8MB
--conky ram 75.6M -62%. swap 14.1M -2%

****Adding iceweasel (took about 60 secs to load):
ID: swap-1 size: 0.42GB used: 0.00GB (0%)
ID: swap-2 size: 0.06GB used: 0.03GB (45%)
ID: swap-3 size: 0.06GB used: 0.03GB (45%)
Memory: 87.7/120.8MB
--conky ram 88.7M -73%, swap 48.0M -9%

****Adding gnumeric:
ID: swap-1 size: 0.42GB used: 0.00GB (0%)
ID: swap-2 size: 0.06GB used: 0.03GB (53%)
ID: swap-3 size: 0.06GB used: 0.03GB (53%)
Memory: 91.5/120.8MB
--conky ram 88.7M -73%, swap 58.3M -11%

****In gnumeric, accessing help F1 to read the gnumeric-doc manual, loads up yelp (65 secs):
ID: swap-1 size: 0.42GB used: 0.02GB (5%)
ID: swap-2 size: 0.06GB used: 0.06GB (100%)
ID: swap-3 size: 0.06GB used: 0.06GB (100%)
Memory: 94.6/120.8MB
--conky ram 98.0M -81%, swap 141M -27%

====> swap -1 kicked in with swap -2/-3 to 100%

****Adding pysolfc (took 45 secs, things are getting slower to load):
ID: swap-1 size: 0.42GB used: 0.07GB (16%)
ID: swap-2 size: 0.06GB used: 0.06GB (100%)
ID: swap-3 size: 0.06GB used: 0.06GB (100%)
Memory: 94.6/120.8MB
--conky ram 97.1M -80%, swap 181M -35%

====> pysolfc is a bit of a ”ram-hog”, but note what happened here. It or something else seems to have been stuffed into swap. Odd. Game does play.

****Adding osmo (30 secs, slow to load):
ID: swap-1 size: 0.42GB used: 0.08GB (20%)
ID: swap-2 size: 0.06GB used: 0.06GB (100%)
ID: swap-3 size: 0.06GB used: 0.06GB (100%)
Memory: 91.9/120.8MB
--conky ram 91.5M -75%, swap 190M -37%

====> note:
-the copy/paste and editing performance within abiword is still fine ie no delay that I can perceive so far, unless I type really fast or do a backspace delete of text, which do go slowly
-abiword menu response to mouse or keyboard not affected, is acceptable
-switching workspaces has gotten slower, hm re-writing the screen each time? Am using all 5.

****Adding synaptic:
ID: swap-1 size: 0.42GB used: 0.10GB (24%)
ID: swap-2 size: 0.06GB used: 0.06GB (100%)
ID: swap-3 size: 0.06GB used: 0.06GB (100%)
Memory: 90.4/120.8MB
--conky ram 94.5M -78%, swap 206M -40%
===> with long idle (coffee break), settled down to 89.1M - 73%, swap 210M -41%


****Plugged in 8 GB USB flashdrive, about 6GB of files. Spacefm, with show thumbnails option on, mounted it, but now we’ve got some problems. Not a total freeze-up, but:
-switching workspaces, I use all 5, very sluggish, need to use keyboard to make this work
-conky disappeared from the screen
-cannot open fluxbox menu with mouse or hotkey
-same with abiword menu, so cannot save the file, but typing in text works OK
-no response to mouse-clicks anywhere
-cannot move the open window with mouse cursor


I’m guessing that the file manager is overwhelmed by the size of the USB flashdrive, plus thumbnails. But I can’t unmount properly, so yanked out the flashdrive. Not good to do that, but it survived. No change, still got those problems. Trying to bail out, I used htop in a virtual console to kill one by one the loaded programmes. First spacefm, no change. Iceweasel, pysofc, no improvement. Yelp, gnumeric, osmo, geany, Control Center, conky. Nope.
What is doing this? Inxi, htop and top show ram/swap demands are down, so they have been released from whatever. Cpu is in the high 90's, but that’s been usual so far. Restarted conky from roxterm, ram 83.7M -65%, swap 53.1M -10%. Still cannot access abiword menu, so killed it to with htop. Still no improvement with rox-fluxbox. Gave up, reboot.
After fresh bootup, rox-fluxbox worked normally, and neither spacefm nor 'Mount Connected Devices”/roxfiler had any problems with the same flashdrive.
Posts: 162
drg
Joined: 22 Feb 2010
#99
Same test on K6-266MHz, but with 256M ram. Overall performance the same with slow cpu/graphics, but why not see if there is a difference? Read into it what you will.

Swap partition still 400MB, inxi sees as 0.42GB, conky as 622M. Zram “Creating 2 devices", ie 2 swapspaces, 114128KiB each.
****Startup with rox-fluxbox
-conky ram 32.7M/248M -13%, swap 0B/622M -0%

****Adding roxterm:
ID: swap-1 size: 0.42GB used: 0.00GB (0%)
ID: swap-2 size: 0.12GB used: 0.00GB (0%)
ID: swap-3 size: 0.12GB used: 0.00GB (0%)
Memory: 40.2/247.7MB
--conky ram 38.2M -15%, swap 0 -0%

****Adding abiword:
ID: swap-1 size: 0.42GB used: 0.00GB (0%)
ID: swap-2 size: 0.12GB used: 0.00GB (0%)
ID: swap-3 size: 0.12GB used: 0.00GB (0%)
Memory: 50.7/247.7MB
--conky ram 47.7M -19%, swap 0B -0%

****Adding Control Center/Edit fluxbox settings, gives geany to review fluxbox hotkeys:
ID: swap-1 size: 0.42GB used: 0.00GB (0%)
ID: swap-2 size: 0.12GB used: 0.00GB (0%)
ID: swap-3 size: 0.12GB used: 0.00GB (0%)
Memory: 64.2/247.7MB
--conky ram 62.1M -25%, swap 0B -0%

****Adding spacefm:
ID: swap-1 size: 0.42GB used: 0.00GB (0%)
ID: swap-2 size: 0.12GB used: 0.00GB (0%)
ID: swap-3 size: 0.12GB used: 0.00GB (0%)
Memory: 80.8/247.7MB
--conky ram 77.9M -31%, swap 0B -0%

****Adding iceweasel (took 30 secs):
ID: swap-1 size: 0.42GB used: 0.00GB (0%)
ID: swap-2 size: 0.12GB used: 0.00GB (0%)
ID: swap-3 size: 0.12GB used: 0.00GB (0%)
Memory: 115.8/247.7MB
--conky ram 113M -45%, swap 96.0K -0%
====> took 30 secs to load, compared to 60 secs with 128M ram (but that time swapping had already started) Here swapping has started barely at 96.0K, if at all.

****Adding gnumeric:
ID: swap-1 size: 0.42GB used: 0.00GB (0%) fs: swap
ID: swap-2 size: 0.12GB used: 0.00GB (0%) fs: swap
ID: swap-3 size: 0.12GB used: 0.00GB (0%) fs: swap
Memory: 127.2/247.7MB
--conky ram 125M -50%, swap 100K -0%


****In gnumeric help (F1) for the manual, invokes yelp:
====> Yikes, same problems as before, but earlier! Mouse cursor does not change to an 'arrow shape, mouse clickings don’t work. Cannot change workspaces now as well.
inxi in console Swap-1 0 -0%, swap-2/-3 0.01GB -5%
conky ram 166M, swap 11.7-1% Swapping in earnest.
Tried to bail out by htop killing, didn’t help. Reboot. Retry.

=====================================================================

****Retry, loading same apps up to gnumeric, but not iceweasel yet, avoiding gnumeric help/yelp:
ID: swap-1 size: 0.42GB used: 0.00GB (0%)
ID: swap-2 size: 0.12GB used: 0.00GB (0%)
ID: swap-3 size: 0.12GB used: 0.00GB (0%)
Memory: 83.6/247.7MB
--conky ram 82.6M -33%, swap 0B -0%

****Adding pysolfc:
ID: swap-1 size: 0.42GB used: 0.00GB (0%)
ID: swap-2 size: 0.12GB used: 0.00GB (0%)
ID: swap-3 size: 0.12GB used: 0.00GB (0%)
Memory: 123.1/247.7MB
--conky ram 129M -48%, swap 36.0K -0%

*****Now adding iceweasel:
ID: swap-1 size: 0.42GB used: 0.00GB (0%)
ID: swap-2 size: 0.12GB used: 0.00GB (4%)
ID: swap-3 size: 0.12GB used: 0.00GB (4%)
Memory: 146.0/247.7MB
--conky ram 143M -57%, swap 8.20M -1%

====> swap kicked in

****Adding osmo:
ID: swap-1 size: 0.42GB used: 0.00GB (0%)
ID: swap-2 size: 0.12GB used: 0.01GB (5%)
ID: swap-3 size: 0.12GB used: 0.01GB (5%)
Memory: 152.4/247.7MB
--conky ram 149M -60%, swap 10.4M -1%

****Adding synaptic:
ID: swap-1 size: 0.42GB used: 0.00GB (0%)
ID: swap-2 size: 0.12GB used: 0.01GB (6%)
ID: swap-3 size: 0.12GB used: 0.01GB (6%)
Memory: 157.0/247.7MB
--conky ram 155M -62%, swap 14.3M -2%

****Now what messed up things last time, plugging in USB thumbdrive with 6Gig of files, with spacefm loaded:
ID: swap-1 size: 0.42GB used: 0.00GB (0%)
ID: swap-2 size: 0.12GB used: 0.01GB (9%)
ID: swap-3 size: 0.12GB used: 0.01GB (9%)
Memory: 171.4/247.7MB
--conky ram 170M -68%, swap 19.0M -3%
====>spacefm mounted USB thumbdrive and opened a directory. Thumbnails option is on, but they don’t show eg a mini-picture of a jpeg.

****Adding roxfiler:
ID: swap-1 size: 0.42GB used: 0.00GB (0%)
ID: swap-2 size: 0.12GB used: 0.02GB (14%)
ID: swap-3 size: 0.12GB used: 0.02GB (14%)
Memory: 177.8/247.7MB
--conky ram 175M -70%, swap 31.8M -5%
=====> Can open and view directories on that USB thumbdrive, thumbnails feature is on and can view mini-picture of jpegs, etc

****Adding Libreoffice writer:
ID: swap-1 size: 0.42GB used: 0.00GB (0%)
ID: swap-2 size: 0.12GB used: 0.04GB (36%)
ID: swap-3 size: 0.12GB used: 0.04GB (36%)
Memory: 169.0/247.7MB
--conky ram 166M -67%, swap 79.8M -12%
====>Can open documents ok, is slower to load than abiword, but does work, just slower . Note, on this box, both are slow.

=====> Overall comments so far. Note that most of these programmes are loaded but not taking any or much cpu time or ram. Some, like LOwriter, pysolfc, osmo and iceweasel take and hold onto ram just idling there. As htop shows “mem% “, LOwriter at 19%, and abiword at 7.3%, but I am using it. At this point, there is a noticeable lag in response to eg typing this text or getting the mouse cursor show it is ready to do something (usually turn into arrow shape here).

****OK, plunge time. Adding gnumeric help/yelp to see if duplicates the partial freeze-ups of last time. First unmount and unplug that thumbdrive, using the antiX feature “unplug USB”:

ID: swap-1 size: 0.42GB used: 0.00GB (0%)
ID: swap-2 size: 0.12GB used: 0.08GB (65%)
ID: swap-3 size: 0.12GB used: 0.08GB (65%)
Memory: 198.3/247.7MB
--conky ram 201M -81%, swap 134M -21%
--conky settled down to ram 195M -78%, swap 146M -23%
====> Well, well. worked OK this time.
Posts: 162
drg
Joined: 22 Feb 2010
#100
antiX-base Edelweißpiraten 26 May 2012
400M swap partition, zram on

K6-266MHz, 128M ram
****At startup Jwm only:
--inxi from console gives swap-1/-2/-3 all at 0 -0%
Memory: 29.6/120.8MB
--conky ram 25.0M/121M -21%, swap 0B/508M -0%

***Adding roxterm:
inxi swap-1/-2/-3 all used 0.00GB 0%
Memory: 34.7/120.8MB
--conky ram 30.5M -25%, swap -0%

****Adding abiword:
inxi swap-1/-2/-3 all used 0.00GB 0%
Memory: 45.2/120.8MB
-conky ram 43.3M -35%, swap 460K/508M -0%

=========> reboot for fresh start, jwm plus roxterm, (but not abiword) then:

****Adding Lowriter:
ID: swap-1 size: 0.42GB used: 0.00GB (0%)
ID: swap-2 size: 0.06GB used: 0.01GB (9%)
ID: swap-3 size: 0.06GB used: 0.01GB (9%)
Memory: 55.9/120.8M
--conky ram 52.9M -43%, swap 9.71M -1%

====> Note, Lowriter takes more ram but not too bad. However, swapping is used already, with the 128M ram.

-----------------------------------------------------------------
-----------------------------------------------------------------

Now with 256M ram

****Jwm, with roxterm:
inxi swap-1/-2/-3 all used 0.00GB 0%
Memory: 35.1/247.7MB
--conky 33.1M -13%, swap -0

****Adding abiword:
inxi swap-1/-2/-3 all used 0.00GB 0%
Memory: 45.8/247.7MB
--conky ram 43.6M -17%, swap 0

====>Reboot for a clean start, jwm plus roxterm, but not abiword, then:

****Adding Lowriter:
ID: swap-1 size: 0.42GB used: 0.00GB (0%)
ID: swap-2 size: 0.12GB used: 0.00GB (0%)
ID: swap-3 size: 0.12GB used: 0.00GB (0%)
Memory: 55.3/247.7MB
--conky ram 59.4M -23%, swap 0 -0%

====> Lowriter takes longer to load (60 secs), but I find the responsiveness of eg menus and nimbleness of adding text to be same or a bit better than abiword. These versions I installed anyway. That's with this 266MHz machine, 256Mram and pretty much without competition for ram and cpu.Swapping isn't used yet with the 256M ram.