Date: Sun, 23 May 2004 23:12:49 +0200 From: Joan Picanyol <lists-freebsd@biaix.org> To: freebsd-questions@freebsd.org Subject: bad gnome responsiveness with network load Message-ID: <20040523211249.GA19381@grummit.biaix.org>
next in thread | raw e-mail | index | archive | help
--pWyiEgJYm5f9v55/
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Hi,
Summary: with /home NFS mounted at 10Mbit/s, network i/o
makes my sytem noticeably slower (as in GNOME interactivity). I'm running
RELENG_5_2, with SCHED_ULE and ADAPTIVE_MUTEXES.
[look at the end for the -questions]
Some system details (long lines):
528,p4,0$ dmesg | grep -v Preloaded | head -16
Copyright (c) 1992-2004 The FreeBSD Project.
Copyright (c) 1979, 1980, 1983, 1986, 1988, 1989, 1991, 1992, 1993, 1994
The Regents of the University of California. All rights reserved.
FreeBSD 5.2.1-RELEASE #0: Sun Apr 11 15:23:13 CEST 2004
joan@calvin.biaix.org:/usr/obj/usr/src/sys/CALVIN
module_register: module vesa already exists!
Module vesa failed to register: 17
Timecounter "i8254" frequency 1193182 Hz quality 0
CPU: AMD Athlon(tm) MP 2400+ (2000.08-MHz 686-class CPU)
Origin = "AuthenticAMD" Id = 0x681 Stepping = 1
Features=0x383fbff<FPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,MMX,FXSR,SSE>
AMD Features=0xc0480000<MP,AMIE,DSP,3DNow!>
real memory = 536346624 (511 MB)
avail memory = 515440640 (491 MB)
ACPI APIC Table: <PTLTD APIC >
FreeBSD/SMP: Multiprocessor System Detected: 2 CPUs
521,p4,0$ cat /etc/libmap.conf
libc_r.so.5 libkse.so.1
libc_r.so libkse.so
I've started to use pan (the GNOME newsreader), and when I "Get New
Headers" it slows the system down: I get long delays when trying to
start gnome-terminal, vim, ssh -X host 'xterm -e mutt', switching
desktops or windows... During this time I see the network load go up
(output of 'netstat -w 5' and 'vmstat -w 5' attached), which I assume
it's because it does heavy i/o from ~/.pan/ (NFS mounted). This doesn't
happen if the system is under high cpu load, and neither does the cpu
load go up when it happens.
Given it's a fast machine and that pan uses at least 4 threads, I wouldn't
expect that one program waiting for the network slows everything down.
Is there anything I might have misconfigured? What else should I look at
to find out what's happening?
tks
--
pica
--pWyiEgJYm5f9v55/
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment; filename="netstat-i5.out"
input (Total) output
packets errs bytes packets errs bytes colls
14 0 2252 14 0 2412 0
6 0 972 6 0 1020 0
4 0 648 4 0 680 0
3 0 486 3 0 510 0
7 0 1158 7 0 1214 0
7 0 1042 7 0 1114 0
51 0 5282 51 0 6266 0
3 0 486 3 0 510 0
3 0 486 3 0 510 0
11 0 1774 11 0 1886 0
11 0 1750 11 0 1934 0
1124 0 1454225 298 0 57734 0
4339 0 5672230 1057 0 188616 0
4222 0 5718896 891 0 156866 0
4119 0 5713968 782 0 142022 0
1070 0 1409540 604 0 601633 0
310 0 62440 2674 0 3795940 0
293 0 59298 2334 0 3294460 0
503 0 109436 3410 0 4740849 0
360 0 90464 2534 0 3510504 0
input (Total) output
packets errs bytes packets errs bytes colls
486 0 93290 2887 0 3920730 0
494 0 110604 3009 0 4133536 0
286 0 57404 2320 0 3262816 0
223 0 45536 1190 0 1568384 0
0 0 0 0 0 0 0
6 0 972 6 0 1020 0
3 0 486 3 0 510 0
3 0 486 3 0 510 0
96 0 27776 57 0 4634 0
122 0 37352 82 0 7824 0
49 0 9330 28 0 2320 0
4 0 546 3 0 510 0
3 0 486 3 0 510 0
7 0 1134 7 0 1190 0
34 0 12756 24 0 2376 0
15 0 2324 18 0 1907 0
--pWyiEgJYm5f9v55/
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment; filename="vmstat-w5.out"
procs memory page disks faults cpu
r b w avm fre flt re pi po fr sr ad0 ad2 in sy cs us sy id
0 3 0 687424 23524 173 0 0 0 165 22 0 0 382 0 701 3 1 96
0 3 0 684448 25048 63 0 0 0 132 0 0 0 456 0 2175 5 2 92
0 3 0 685572 24864 20 0 0 0 12 0 0 1 341 0 815 2 2 96
0 3 0 685572 24864 0 0 0 0 0 0 0 0 339 0 532 1 1 98
0 3 0 686680 24704 20 0 0 0 12 0 0 1 340 0 582 2 2 97
0 3 0 686712 24672 25 0 0 0 24 0 0 1 365 0 636 2 1 97
0 3 1 687820 24512 16 0 0 0 9 0 0 1 341 0 721 2 1 97
1 3 1 687820 24508 179 0 0 0 162 0 0 0 362 0 540 1 2 97
0 3 0 688924 24344 20 0 0 0 7 0 0 1 342 0 906 2 2 96
0 3 0 688928 24340 4 0 0 0 4 0 0 0 374 0 743 3 2 95
0 3 0 688928 24340 3 0 0 0 3 0 0 2 395 0 813 2 2 96
0 3 0 688928 24340 88 0 0 0 88 0 0 0 443 0 2257 5 3 92
0 4 0 688928 40936 103 0 0 0 1026 0 0 3 1116 0 4204 7 4 89
0 5 0 691184 33688 652 0 0 0 371 0 0 0 2369 0 8276 13 7 80
0 4 1 692276 28512 220 0 0 0 124 0 0 1 2200 0 6755 7 7 87
0 4 0 690956 23200 553 0 0 0 316 0 0 0 2262 0 8075 13 8 79
0 4 0 696272 31672 590 7 10 0 780 738 0 16 859 0 3410 27 4 68
0 6 0 692240 30272 110 1 1 0 141 0 0 10 1073 0 4432 4 5 91
procs memory page disks faults cpu
r b w avm fre flt re pi po fr sr ad0 ad2 in sy cs us sy id
0 5 1 692240 28040 24 0 0 0 39 0 0 2 967 0 3654 3 5 93
0 6 0 692240 24896 3 0 0 0 159 0 0 10 1214 0 5034 2 5 92
0 4 0 692240 23072 32 1 1 0 137 0 0 2 982 0 3609 3 4 93
0 5 0 692372 19976 26 0 0 0 154 0 0 2 1090 0 4276 1 5 94
0 4 1 692500 29840 7 0 0 1 137 4300 0 1 1089 0 4096 1 5 94
0 4 0 692500 27784 1 0 0 0 103 0 0 2 887 0 2923 0 4 96
0 3 0 692500 27468 0 0 0 0 16 0 0 0 584 0 1730 3 2 95
0 3 0 692500 27468 0 0 0 0 0 0 0 0 331 0 363 0 1 99
0 3 1 692500 27468 0 0 0 0 0 0 0 1 332 0 368 0 1 99
0 3 0 692500 27468 0 0 0 0 0 0 0 0 331 0 366 0 1 99
0 3 0 692500 27468 0 0 0 0 0 0 0 0 331 0 361 0 1 98
0 3 0 692500 27464 0 0 0 0 0 0 0 0 386 0 584 0 2 98
1 3 0 692500 27464 1 0 0 0 0 0 0 0 393 0 689 0 1 98
0 3 0 692500 27464 1 0 0 0 0 0 0 0 355 0 443 0 1 99
0 3 0 692500 27464 0 0 0 0 1 0 0 0 332 0 355 0 1 99
0 3 0 692500 27464 0 0 0 0 0 0 0 1 335 0 356 0 1 99
1 3 0 692500 27464 0 0 0 0 0 0 0 0 331 0 331 0 1 99
0 3 1 692500 27464 0 0 0 0 1 0 0 2 354 0 535 0 2 98
procs memory page disks faults cpu
r b w avm fre flt re pi po fr sr ad0 ad2 in sy cs us sy id
1 3 0 689648 28456 37 0 0 0 76 0 0 0 696 0 2226 6 3 92
--pWyiEgJYm5f9v55/--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20040523211249.GA19381>