Return to BSD News archive
Path: euryale.cc.adfa.oz.au!newshost.carno.net.au!harbinger.cc.monash.edu.au!news.cs.su.oz.au!news.cs.su.oz.au!inferno.mpx.com.au!news.ci.com.au!brian.telstra.net!act.news.telstra.net!vic.news.telstra.net!news.mira.net.au!pumpkin.pangea.ca!www.nntp.primenet.com!nntp.primenet.com!news.mathworks.com!cam-news-hub1.bbnplanet.com!atl-news-feed1.bbnplanet.com!news.bbnplanet.com!futuresouth.com!tim
From: tim@futuresouth.com (Tim Tsai)
Newsgroups: comp.unix.bsd.freebsd.misc
Subject: help improve news server performance
Date: 1 Jul 1997 08:03:05 GMT
Organization: FutureSouth Communications, Inc.
Lines: 170
Message-ID: <5padjp$1ot$1@news.futuresouth.com>
NNTP-Posting-Host: shell.futuresouth.com
X-Newsreader: NN version 6.5.1 (NOV)
Xref: euryale.cc.adfa.oz.au comp.unix.bsd.freebsd.misc:43759
I have set up what I thought is a reasonable news server but apparently
it's not keeping up with our single feed (we have a minor local feed but
it's like 20 articles a day so we can safely ignore it). I have installed
Joe Greco's innd timer patch and am including some statistics. Any ideas?
We did not have this problem when we were fed by another machine (but
quite a bit further network wise) but I want to make sure it's not
something on our end.
Feed:
Approximately 320k articles per day (the big 7 including binaries),
split across 6 streaming feeds. We have played with the # of
feeds and it doesn't seem to help.
Network:
I am 4 hops away from the feed. We're talking about two routers
(ours and theirs) and a clear T1 (25% utilized) here. The ping
time on the feeding machine is about 25ms.
NNRP Readers:
Rarely more than a handful.
Machine:
AMD K5-133 with 128 megs of RAM. Swap is NEVER used.
Software:
INND 1.5.1
FreeBSD 2.2-STABLE (6/30/97)
Spool:
ccd0: two Quantum Atlas II's (Ultra-Wide SCSI) on an Adaptec 2940UW.
It contains only alt.binaries.*
ccd1: five IBM DOR-32160W (Ultra-Wide SCSI) on another Adaptec 2940UW.
It contains the rest of the hierarchies.
These total about 17G's. Both mounted with noatime.
History:
old Fujitsu SCSI-2 drive on an NCR 53c875 controller. While this is
a fairly slow drive as far as transfer rate goes it has good seek
times.
Over.view:
Quantum Fireball on the same NCR 53c875 controller. mounted noatime.
Root/Tmp/Swap:
Quantum Fireball on the same NCR 53c875.
The article times seem to be pretty lousy to me and is possibly the
limiting factor. Any ideas?
Also, is there anyway for me to check that it's not my feed that's
behind?
Thanks!
Tim
-- output from ccd.conf --
ccd0 65536 0 /dev/sd3e /dev/sd6e
ccd1 65536 0 /dev/sd4e /dev/sd5e /dev/sd7e /dev/sd8e /dev/sd9e
-- output from Joe's innd timer patches --
Times are in milliseconds. The format is time(transactions), so to get
the average just divide the first number by the number in parenthesis.
Jul 1 00:35:01 news innd: ME time 300244 idle 145476(1268) artwrite 93221(666) artlink 10(432) hiswrite 22194(704) hissync 265(133) sitesend 44(1332) artctrl 13274(259) artcncl 0(0) hishave 28891(1546) hisgrep 15(33)
Jul 1 00:40:01 news innd: ME time 300042 idle 122412(4094) artwrite 101587(848) artlink 12(428) hiswrite 29979(918) hissync 70(170) sitesend 56(1696) artctrl 9635(177) artcncl 0(0) hishave 34931(1115) hisgrep 16(23)
Jul 1 00:45:03 news innd: ME time 301963 idle 92596(3509) artwrite 110698(1139) artlink 19(981) hiswrite 39995(1276) hissync 134(228) sitesend 79(2278) artctrl 14753(224) artcncl 0(0) hishave 43363(1500) hisgrep 16(58)
Jul 1 00:50:03 news innd: ME time 300522 idle 85127(4464) artwrite 113921(1112) artlink 17(712) hiswrite 37850(1172) hissync 45(223) sitesend 61(2224) artctrl 20937(202) artcncl 0(0) hishave 39709(1373) hisgrep 33(133)
Jul 1 00:55:03 news innd: ME time 300226 idle 75917(3305) artwrite 118336(932) artlink 17(817) hiswrite 32363(996) hissync 328(187) sitesend 42(1864) artctrl 35031(579) artcncl 107(1) hishave 48960(1576) hisgrep 54(126)
Jul 1 01:00:03 news innd: ME time 300074 idle 67721(4029) artwrite 117009(1084) artlink 132(2190) hiswrite 39974(1155) hissync 549(217) sitesend 71(2168) artctrl 21074(201) artcncl 0(1) hishave 45125(1357) hisgrep 35(111)
Jul 1 02:10:02 news innd: ME time 302935 idle 81776(3151) artwrite 165951(690) artlink 18(855) hiswrite 981(705) hissync 34(138) sitesend 79(1380) artctrl 23744(115) artcncl 0(0) hishave 26326(1415) hisgrep 99(65)
Jul 1 02:15:03 news innd: ME time 300199 idle 76939(5409) artwrite 159561(804) artlink 26(798) hiswrite 2505(866) hissync 37(161) sitesend 56(1608) artctrl 22571(130) artcncl 199(2) hishave 33048(1013) hisgrep 39(91)
Jul 1 02:20:03 news innd: ME time 300002 idle 66297(3234) artwrite 137245(826) artlink 60(791) hiswrite 9019(873) hissync 79(165) sitesend 60(1652) artctrl 52703(287) artcncl 1(1) hishave 35665(1161) hisgrep 80(193)
Jul 1 02:25:03 news innd: ME time 300909 idle 60446(3585) artwrite 174600(832) artlink 16(886) hiswrite 15950(867) hissync 366(168) sitesend 99(1664) artctrl 9695(63) artcncl 0(0) hishave 32010(930) hisgrep 37(47)
-- output of dmesg --
Copyright (c) 1992-1997 FreeBSD Inc.
Copyright (c) 1982, 1986, 1989, 1991, 1993
The Regents of the University of California. All rights reserved.
FreeBSD 2.2-STABLE #0: Mon Jun 30 21:02:04 CDT 1997
root@:/usr/src/sys/compile/FSKERNEL
CPU: AMD K5 model 1 (99.72-MHz 586-class CPU)
Origin = "AuthenticAMD" Id = 0x511 Stepping=1
Features=0x21bf<FPU,VME,DE,PSE,TSC,MSR,MCE,CX8,PGE>
real memory = 134217728 (131072K bytes)
avail memory = 129544192 (126508K bytes)
Probing for devices on PCI bus 0:
chip0 <Intel 82439> rev 1 on pci0:0
chip1 <Intel 82371SB PCI-ISA bridge> rev 1 on pci0:7:0
chip2 <Intel 82371SB IDE interface> rev 0 on pci0:7:1
ncr0 <ncr 53c875 wide scsi> rev 1 int a irq 9 on pci0:9
(ncr0:0:0): "QUANTUM FIREBALL_TM2110S 300N" type 0 fixed SCSI 2
sd0(ncr0:0:0): Direct-Access
sd0(ncr0:0:0): 10.0 MB/s (100 ns, offset 15)
2014MB (4124736 512 byte sectors)
sd0(ncr0:0:0): with 6810 cyls, 4 heads, and an average 151 sectors/track
(ncr0:1:0): 4.4 MB/s (225 ns, offset 15)
(ncr0:1:0): "FUJITSU M2623S-512 0303" type 0 fixed SCSI 2
sd1(ncr0:1:0): Direct-Access
sd1(ncr0:1:0): 5.0 MB/s (200 ns, offset 15)
405MB (830340 512 byte sectors)
sd1(ncr0:1:0): with 1429 cyls, 9 heads, and an average 64 sectors/track
(ncr0:4:0): "QUANTUM FIREBALL_TM2110S 300N" type 0 fixed SCSI 2
sd2(ncr0:4:0): Direct-Access
sd2(ncr0:4:0): 10.0 MB/s (100 ns, offset 15)
2014MB (4124736 512 byte sectors)
sd2(ncr0:4:0): with 6810 cyls, 4 heads, and an average 151 sectors/track
ahc0 <Adaptec 2940 Ultra SCSI host adapter> rev 0 int a irq 12 on pci0:11
ahc0: aic7880 Wide Channel, SCSI Id=7, 16/255 SCBs
ahc0: target 1 Tagged Queuing Device
(ahc0:1:0): "QUANTUM XP34550W LXQ1" type 0 fixed SCSI 2
sd3(ahc0:1:0): Direct-Access 4341MB (8890760 512 byte sectors)
sd3(ahc0:1:0): with 5899 cyls, 10 heads, and an average 150 sectors/track
Sending SDTR!!
ahc0: target 4 Tagged Queuing Device
(ahc0:4:0): "IBM DORS-32160W !# WA3E" type 0 fixed SCSI 2
sd4(ahc0:4:0): Direct-Access 2063MB (4226725 512 byte sectors)
sd4(ahc0:4:0): with 6703 cyls, 5 heads, and an average 126 sectors/track
Sending SDTR!!
ahc0: target 8 Tagged Queuing Device
(ahc0:8:0): "IBM DORS-32160W !# WA3E" type 0 fixed SCSI 2
sd5(ahc0:8:0): Direct-Access 2063MB (4226725 512 byte sectors)
sd5(ahc0:8:0): with 6703 cyls, 5 heads, and an average 126 sectors/track
ahc1 <Adaptec 2940 Ultra SCSI host adapter> rev 0 int a irq 11 on pci0:12
ahc1: aic7880 Wide Channel, SCSI Id=7, 16/255 SCBs
ahc1: target 1 Tagged Queuing Device
(ahc1:1:0): "QUANTUM XP34550W LXQ1" type 0 fixed SCSI 2
sd6(ahc1:1:0): Direct-Access 4341MB (8890760 512 byte sectors)
sd6(ahc1:1:0): with 5899 cyls, 10 heads, and an average 150 sectors/track
Sending SDTR!!
ahc1: target 2 Tagged Queuing Device
(ahc1:2:0): "IBM DORS-32160W !# WA3E" type 0 fixed SCSI 2
sd7(ahc1:2:0): Direct-Access 2063MB (4226725 512 byte sectors)
sd7(ahc1:2:0): with 6703 cyls, 5 heads, and an average 126 sectors/track
Sending SDTR!!
ahc1: target 4 Tagged Queuing Device
(ahc1:4:0): "IBM DORS-32160W !# WA3E" type 0 fixed SCSI 2
sd8(ahc1:4:0): Direct-Access 2063MB (4226725 512 byte sectors)
sd8(ahc1:4:0): with 6703 cyls, 5 heads, and an average 126 sectors/track
Sending SDTR!!
ahc1: target 8 Tagged Queuing Device
(ahc1:8:0): "IBM DORS-32160W !# WA3E" type 0 fixed SCSI 2
sd9(ahc1:8:0): Direct-Access 2063MB (4226725 512 byte sectors)
sd9(ahc1:8:0): with 6703 cyls, 5 heads, and an average 126 sectors/track
Probing for devices on the ISA bus:
sc0 at 0x60-0x6f irq 1 on motherboard
sc0: VGA color <16 virtual consoles, flags=0x0>
ed0 at 0x300-0x31f irq 10 on isa
ed0: address 00:40:33:90:cd:16, type NE2000 (16 bit)
sio0 at 0x3f8-0x3ff irq 4 on isa
sio0: type 16550A
sio1 at 0x2f8-0x2ff irq 3 on isa
sio1: type 16550A
sio2: disabled, not probed.
sio3: disabled, not probed.
lpt0 at 0x378-0x37f irq 7 on isa
lpt0: Interrupt-driven port
lp0: TCP/IP capable interface
fdc0 at 0x3f0-0x3f7 irq 6 drq 2 on isa
fdc0: NEC 72065B
fd0: 1.44MB 3.5in
wdc0 not found at 0x1f0
wdc1 not found at 0x170
npx0 on motherboard
npx0: INT 16 interface
apm0: disabled, not probed.
changing root device to sd0a
ccd0-3: Concatenated disk drivers