Date: Fri, 2 Oct 2009 10:17:24 -0700 (PDT) From: Richard Mahlerwein <mahlerrd@yahoo.com> To: freebsd-performance@freebsd.org Subject: Re: A specific example of a disk i/o problem Message-ID: <969908.40235.qm@web51006.mail.re2.yahoo.com>
next in thread | raw e-mail | index | archive | help
>From: Dieter <freebsd@sopwith.solgatos.com>=0A>Subject: Re: A specific exa= mple of a disk i/o problem=0A>To: freebsd-performance@freebsd.org=0A>Date: = Friday, October 2, 2009, 1:07 AM=0A>=0A>Updated demo, just to make sure:=0A= >=0A># big_file is larger than main memory, on same disk as man (/usr)=0A>t= ime man de=A0=A0=A0 =A0=A0=A0 # get baseline time for man command without >= competing i/o=0A>cat big_file > /dev/null=A0=A0=A0# flush man command and d= ata from memory=0A>cat big_file > /dev/null &=A0# generate i/o, attempt to = use up bottleneck=0A>time man de=A0 =A0=A0=A0 # see how much longer man tak= es with competing >i/o=0A=0ATrying it on something that surely will have i/= o issues: PIII 500 w/ 60 GB PATA drive and 128 MB RAM. Blank minimal insta= ll + man pages right now. I'm trying with a "file /etc/" command instead o= f trying to reliably time man... It seems the file command always hits the = disk. This should also make any sort of "flushing" unecessary, shouldn't i= t?=0A=0Aliebnitz# uname -a=0AFreeBSD xyz 7.2-RELEASE FreeBSD 7.2-RELEASE #0= : Fri May 1 08:49:13 UTC 2009 root@walker.cse.buffalo.edu:/usr/obj/usr= /src/sys/GENERIC i386=0A=0Aliebnitz# time dd if=3D/dev/zero of=3D/usr/bigf= ile count=3D4096 bs=3D1M=0A4096+0 records in=0A4096+0 records out=0A4294967= 296 bytes transferred in 363.700967 secs (11809062 bytes/sec)=0A0.055u 83.9= 71s 6:03.78 23.0% 25+1071k 128+34391io 0pf+0w=0A=0Aliebnitz# time file /= etc/ [run 5 times]=0A4.052u 0.047s 0:04.14 98.7% 15+1085k 0+0io 0pf+0w= =0A4.068u 0.031s 0:04.13 99.0% 16+1095k 0+0io 0pf+0w=0A4.052u 0.046s 0:= 04.13 99.0% 16+1101k 0+0io 0pf+0w=0A4.044u 0.054s 0:04.13 99.0% 16+= 1092k 0+0io 0pf+0w=0A4.052u 0.046s 0:04.13 99.0% 15+1091k 0+0io 0pf+0w= =0A=0AOk, so that looks pretty repeatable. Now...=0Aliebnitz# cat /usr/big= file > /dev/null &=0A[1] 37923=0Aliebnitz# time file /etc/* [again run 5 ti= mes]=0A4.146u 0.038s 0:06.80 61.3% 15+1103k 87+0io 8pf+0w=0A4.128u 0.05= 4s 0:06.73 61.9% 15+1095k 93+0io 6pf+0w=0A =0A=0A=0A=0A=0AFlush as exam= ple.=0Aliebnitz# cat /usr/bigfile > /dev/null=0Aliebnitz# cat /usr/bigfile = > /dev/null &=0A[1] 37890=0A=0A=0A=0A
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?969908.40235.qm>