Date: Tue, 17 Jan 1995 17:49:14 +0800 (SST) From: SysAdmin - Ng Pheng Siong <lsys@np.ac.sg> To: hackers@FreeBSD.org Cc: lsys@np.ac.sg (SysAdmin) Subject: Re: Filesystem(?) preformance - 1.x and 2.0 Message-ID: <199501170950.BAA03993@freefall.cdrom.com> In-Reply-To: <199501111752.SAA01305@knobel.gun.de> from "Andreas Klemm" at Jan 11, 95 06:52:55 pm
next in thread | previous in thread | raw e-mail | index | archive | help
Andreas wrote:
> My comment: "Check cables and connectors" ;-)
Checked.
Well, I've just discovered the '-l' option to 'time'. ;)
Here are 2 consecutive runs of `time -l cat web2 > /dev/null' on the 1.x:
/home/ngps/perf:$ time -l cat web2 >/dev/null
5.07 real 0.02 user 0.56 sys
160 maximum resident set size
34 average shared memory size
19 average unshared data size
128 average unshared stack size
204 page reclaims
0 page faults
0 swaps
306 block input operations
0 block output operations
0 messages sent
0 messages received
0 signals received
305 voluntary context switches
6 involuntary context switches
/home/ngps/perf:$ time -l cat web2 >/dev/null
0.36 real 0.02 user 0.34 sys
152 maximum resident set size
34 average shared memory size
18 average unshared data size
128 average unshared stack size
48 page reclaims
0 page faults
0 swaps
0 block input operations
0 block output operations
0 messages sent
0 messages received
0 signals received
0 voluntary context switches
4 involuntary context switches
The 2nd time around, no block input operation!
On the 2.0 system, each of 2 consecutive runs causes block input ops.
Can someone kindly provide a pointer as to what's happening, please?
- PS
--
Ng Pheng Siong * lsys@np.ac.sg * ngps@np.ac.sg
Computer Centre, Ngee Ann Polytechnic, Singapore
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199501170950.BAA03993>