gjalt@ele.tue.nl (Gjalt de Jong) (02/14/91)
In addition to previous postings on process sizes:
Who can explain to me why a 'ps l' and a 'ps u' do not agree on the sizes of
the processes. For example:
ps aux:
USER PID SZ RSS TTY STAT TIME COMMAND
root 1 768 536 ? S < 0:12 /etc/init
root 2 0 0 ? R 1207:17 null
root 3 0 0 ? S 1:01 purifier
root 4 0 0 ? S 0:02 purifier
root 5 0 0 ? S 2:24 unwired_dxm
root 6 0 0 ? S 0:00 pinger
root 7 0 0 ? S 0:42 netreceive
root 8 0 0 ? S 0:39 netpaging
root 9 0 0 ? S 1:39 wired_dxm
root 10 0 0 ? S 2:11 netrequest
root 89 512 124 ? S 8:11 /etc/tcpd -c
root 96 512 72 ? S 0:04 /etc/syslogd
root 99 512 68 ? S 0:20 /etc/routed -f -q
root 102 512 64 ? S 0:00 /etc/inetd
root 105 512 140 ? S 0:01 /etc/named
root 109 256 64 ? S 0:31 /etc/rwhod
root 113 256 124 ? S < 1:45 /usr/local/bin/xntpd -b
root 116 512 124 ? S 0:00 /etc/snmpd
root 121 256 68 ? S 0:02 /etc/ncs/llbd
root 125 256 64 ? S 0:43 /etc/cron
user 133 256 44 ? S 0:04 /sys/spm/spm
user 135 256 76 ? S 0:00 /etc/writed
user 139 512 12 ? S 0:00 /sys/mbx/mbx_helper
root 142 2048 360 ? S < 19:53 /etc/Xapollo -K /usr/X11/lib/keyboard/keyboard.config -D1 s+r-
sys_pers 145 512 824 ? S < 9:39 dm
root 748 512 168 ? S 0:00 rshd
michel 754 1792 600 ? S 3:39 gnuemacs -i -font f7x13.b -w =80x55-0+0 -b 3 -ms red
root 1356 512 40 ? S 0:25 rlogind
root 1430 512 52 ? S 0:04 rlogind
root 1591 512 172 ? S 0:00 rshd
gjalt 1592 512 192 ? S 0:00 csh -c ps aux
gjalt 1596 512 172 ? R 0:00 ps aux
geert 1357 512 140 ttyp0 S 0:04 -sh
geert 1585 256 136 ttyp0 S 0:00 time n_ptl -v -R ../examples/mul4.ptl
geert 1586 5120 4384 ttyp0 R 5:17 n_ptl -v -R ../examples/mul4.ptl
frans 870 256 76 ttyp1 S 0:16 /usr/local/emacs-18.56/etc/loadst -n 60
michel 755 256 76 ttyp2 S 0:20 /usr/local/emacs-18.56/etc/loadst -n 60
michel 756 512 196 ttyp3 S 0:02 /bin/csh -i
michel 1235 512 112 ttyp3 S 0:01 /bin/sh decom_all apex3.log apex4.lo
michel 135510624016680 ttyp3 R 570:12 log_decom primes12
ed 1431 512 132 ttyp4 S 0:02 -csh
ed 1547 256 536 ttyp4 S 0:02 xload
michel 822 512 128 ttyp5 S 0:03 /bin/csh -i
michel 826 256 184 ttyp5 S 1:34 xload
frans 869 1024 424 ttyp5 S 1:08 gnuemacs
frans 871 512 120 ttyp6 S 0:01 /bin/csh -i
ps alx:
F UID PID PPID CP PRI NI ADDR SZ RSS STAT TTY TIME COMMAND
1008001 0 1 1 0 1-20 1 512 536 S < ? 0:12 /etc/init
3 0 2 0 0 17 0 0 0 0 R ? 1207:17 null
3 0 3 0 0 1 0 0 0 0 S ? 1:01 purifier
3 0 4 0 0 1 0 0 0 0 S ? 0:02 purifier
3 0 5 0 0 1 0 0 0 0 S ? 2:25 unwired_dxm
3 0 6 0 0 1 0 0 0 0 S ? 0:00 pinger
3 0 7 0 0 1 0 0 0 0 S ? 0:42 netreceive
3 0 8 0 0 1 0 0 0 0 S ? 0:39 netpaging
3 0 9 0 0 1 0 0 0 0 S ? 1:39 wired_dxm
3 0 10 0 0 1 0 0 0 0 S ? 2:11 netrequest
1008001 0 89 1 0 3 0 5 256 124 S ? 8:11 /etc/tcpd -c
1008001 0 96 1 0 3 0 6 256 72 S ? 0:04 /etc/syslogd
1008001 0 99 1 0 3 0 7 256 68 S ? 0:20 /etc/routed -f -q
1008001 0 102 1 0 3 0 8 256 64 S ? 0:00 /etc/inetd
1008001 0 105 1 0 3 0 9 256 132 S ? 0:01 /etc/named
1008001 0 109 1 0 3 0 a 256 64 S ? 0:31 /etc/rwhod
1008001 0 113 1 0 1-12 c 256 124 S < ? 1:45 /usr/local/bin/xntpd -b
1008001 0 116 1 0 3 0 d 256 124 S ? 0:00 /etc/snmpd
1008001 0 121 1 0 3 0 4 256 68 S ? 0:02 /etc/ncs/llbd
1008001 0 125 1 0 3 0 e 256 68 S ? 0:43 /etc/cron
1008001 14 133 1 0 3 0 10 256 44 S ? 0:04 /sys/spm/spm
1008001 14 135 1 0 3 0 11 256 76 S ? 0:00 /etc/writed
1008001 14 139 1 0 3 0 12 256 12 S ? 0:00 /sys/mbx/mbx_helper
1008001 0 142 1 0 1 -6 3 1536 360 S < ? 19:53 /etc/Xapollo -K /usr/X11/
1008001 13 145 1 0 1-20 2 256 824 S < ? 9:39 dm
1008001 0 748 102 0 3 0 17 256 168 S ? 0:00 rshd
1008001 38 754 1 0 3 0 1d 1792 608 S ? 3:39 gnuemacs -i -font
1008001 0 1356 102 0 3 0 16 256 44 S ? 0:25 rlogind
1008001 0 1430 102 0 3 0 19 256 52 S ? 0:04 rlogind
1008001 0 1597 102 0 3 0 23 256 164 S ? 0:00 rshd
1008001 20 1598 1597 0 5 0 24 256 196 R ? 0:00 csh -c ps alx
1008001 20 1602 1598 0 3 0 25 256 176 R ? 0:00 ps alx
1008001 19 1357 1356 0 3 0 18 256 140 S ttyp0 0:04 -sh
1008001 19 1585 1357 0 3 0 1e 256 136 S ttyp0 0:00 time n_ptl -v -R ../examples/mul4.ptl
1008001 19 1586 1585 0 14 0 20 4864 4040 R ttyp0 5:45 n_ptl -v -R ../examples/mul4.ptl
1008001 18 870 869 0 3 0 13 256 76 S ttyp1 0:16 /usr/local/emacs-18.56/etc/loadst -n 60
1008001 38 755 754 0 3 0 1b 256 80 S ttyp2 0:20 /usr/local/emacs-18.56/etc/loadst -n 60
1008001 38 756 754 0 3 0 1c 256 196 S ttyp3 0:02 /bin/csh -i
1008001 38 1235 756 0 3 0 b 256 112 S ttyp3 0:01 /bin/sh decom_all apex3.log
1008001 38 1355 1235 0 14 0 f10598416676 R ttyp3 570:40 log_decom primes12
1008001 68 1431 1430 0 3 0 1a 256 132 S ttyp4 0:02 -csh
1008001 68 1547 1431 0 3 0 22 256 536 S ttyp4 0:02 xload
1008001 38 822 754 0 3 0 1f 256 128 S ttyp5 0:03 /bin/csh -i
1008001 38 826 822 0 3 0 21 256 184 S ttyp5 1:35 xload
1008001 18 869 1 0 4 0 14 1024 424 S ttyp5 1:08 gnuemacs
1008001 18 871 869 0 3 0 15 256 120 S ttyp6 0:01 /bin/csh -i
Especially note all the small processes with a size of 256 or 512kB
BTW, these ps's were run on a node running 10.3, but it's same for 10.2
Some more weirdness:
a running lisp after running 5 minutes, with a size of 16MB, showed to have a
resident size of only 156kB (on a DN2500 with 16MB memory), and there were
no other time consuming processes running. Amazing isn't it?
--
__
Gjalt G. de Jong, | Phone: +(31)40-473345
Eindhoven University of Technology, Dept. of Electr. Eng. (ES/EH 7.26)
P.O. Box 513, 5600 MB Eindhoven, The Netherlands
Email: gjalt@ele.tue.nlkrowitz@RICHTER.MIT.EDU (David Krowitz) (02/15/91)
It appears that the listing from /bin/ps -alx shows the "size" to be 256 KB less than the listing from /bin/ps -aux. It occurs to me that the default stack size is (if I'm not wrong) 256 KB. Could this be the difference? -- David Krowitz krowitz@richter.mit.edu (18.83.0.109) krowitz%richter.mit.edu@eddie.mit.edu krowitz%richter.mit.edu@mitvma.bitnet (in order of decreasing preference)