Date: 02 Jul 1999 10:02:32 -0400 From: Arcady Genkin <a.genkin@utoronto.ca> To: Alfred Perlstein <bright@rush.net> Cc: Arcady Genkin <a.genkin@utoronto.ca>, freebsd-questions@freebsd.org Subject: Re: FreeBSD is painfully slow on my 486 Message-ID: <87btdvywsn.fsf@main.wgaf.net> In-Reply-To: Alfred Perlstein's message of "Fri, 02 Jul 1999 03:31:32 -0500 (EST)" References: <Pine.BSF.3.96.990702032729.14320p-100000@cygnus.rush.net>
next in thread | previous in thread | raw e-mail | index | archive | help
Alfred Perlstein <bright@rush.net> writes:
> I think you need to give us some perspective of what's wrong:
>
> 1) are cpu bound things slow?
> 2) are programs that access a lot of memory slow when not swapping
> this will test the cpu cache somewhat
> 3) is paging really slow?
> 4) what about I/O? on raw devices, on block devices?
> 5) does top show that you have almost no free memory?
>
> for a 486, i would recommend enabling multi-block and 32 bit
> transfers if possible, leave DMA off, i'm quite sure it wasn't
> available for that kind of hardware.
It seems to me that the things are not only affected by slower
I/O. For example scrolling in /sdand/sysinstall I can feel perceptible
delays between me pressing "down" key and the cursor actually moving
down. Same for other programs. I'm aware that I/O is slower with
synchroniously mounted devices, but I don't think it should cause
system slowing down *that* much.
last pid: 351; load averages: 0.14, 0.13, 0.08 up 0+00:09:15 11:56:17
17 processes: 1 running, 16 sleeping
Mem: 3216K Active, 4828K Inact, 2784K Wired, 1426K Buf, 3752K Free
Swap: 40M Total, 40M Free
PID USERNAME PRI NICE SIZE RES STATE TIME WCPU CPU COMMAND
318 antipode 2 0 1636K 1284K select 0:09 1.78% 1.76% midc
152 root 2 0 888K 620K select 0:02 0.00% 0.00% inetd
309 antipode 10 0 500K 320K wait 0:02 0.00% 0.00% sh
308 root 2 0 876K 636K select 0:02 0.00% 0.00% telnetd
316 antipode 10 0 500K 352K wait 0:02 0.00% 0.00% sh
111 root 2 0 820K 572K select 0:02 0.00% 0.00% syslogd
315 root 2 0 876K 636K select 0:01 0.00% 0.00% telnetd
351 antipode 50 0 1548K 824K RUN 0:01 0.00% 0.00% top
305 root 3 0 824K 572K ttyin 0:01 0.00% 0.00% getty
307 root 3 0 824K 572K ttyin 0:01 0.00% 0.00% getty
306 root 3 0 824K 572K ttyin 0:01 0.00% 0.00% getty
155 root 10 0 980K 592K nanslp 0:00 0.00% 0.00% cron
1 root 10 0 420K 236K wait 0:00 0.00% 0.00% init
159 root 2 0 1304K 952K select 0:00 0.00% 0.00% sendmail
299 root 2 0 480K 288K select 0:00 0.00% 0.00% dhclient
120 daemon 2 0 820K 400K select 0:00 0.00% 0.00% portmap
42 root 18 0 204K 84K pause 0:00 0.00% 0.00% adjkerntz
I'm not sure if it made to the list the last time, so I'll repost
benchmark results for that machine -- can you guys have a look and
tell me where the things are slower than they shoud be for an i486 DX4
with 16M of RAM?
==============================================================
BYTE UNIX Benchmarks (Version 3.11)
System -- door.wgaf.net
Start Benchmark Run: Thu Jul 1 07:32:18 EST 1999
3 interactive users.
Dhrystone 2 without register variables 4713.8 lps (10 secs, 6 samples)
Dhrystone 2 using register variables 6390.7 lps (10 secs, 6 samples)
Arithmetic Test (type = arithoh) 149278.2 lps (10 secs, 6 samples)
Arithmetic Test (type = register) 1746.4 lps (10 secs, 6 samples)
Arithmetic Test (type = short) 3861.1 lps (10 secs, 6 samples)
Arithmetic Test (type = int) 2211.7 lps (10 secs, 6 samples)
Arithmetic Test (type = long) 2211.1 lps (10 secs, 6 samples)
Arithmetic Test (type = float) 5517.4 lps (10 secs, 6 samples)
Arithmetic Test (type = double) 5573.2 lps (10 secs, 6 samples)
System Call Overhead Test 1948.5 lps (10 secs, 6 samples)
Pipe Throughput Test 1624.5 lps (10 secs, 6 samples)
Pipe-based Context Switching Test 757.9 lps (10 secs, 6 samples)
Process Creation Test 72.0 lps (10 secs, 6 samples)
Execl Throughput Test 41.2 lps (10 secs, 6 samples)
File Read (10 seconds) 2908.0 KBps (10 secs, 6 samples)
File Write (10 seconds) 600.0 KBps (10 secs, 6 samples)
File Copy (10 seconds) 360.0 KBps (10 secs, 6 samples)
File Read (30 seconds) 3140.0 KBps (30 secs, 6 samples)
File Write (30 seconds) 533.0 KBps (30 secs, 6 samples)
File Copy (30 seconds) 392.0 KBps (30 secs, 6 samples)
C Compiler Test 7.0 lpm (60 secs, 3 samples)
Shell scripts (1 concurrent) 11.0 lpm (60 secs, 3 samples)
Shell scripts (2 concurrent) 5.3 lpm (60 secs, 3 samples)
Shell scripts (4 concurrent) 2.6 lpm (60 secs, 3 samples)
Shell scripts (8 concurrent) 1.0 lpm (60 secs, 3 samples)
Dc: sqrt(2) to 99 decimal places 1324.3 lpm (60 secs, 6 samples)
Recursion Test--Tower of Hanoi 71.9 lps (10 secs, 6 samples)
INDEX VALUES
TEST BASELINE RESULT INDEX
Arithmetic Test (type = double) 2541.7 5573.2 2.2
Dhrystone 2 without register variables 22366.3 4713.8 0.2
Execl Throughput Test 16.5 41.2 2.5
File Copy (30 seconds) 179.0 392.0 2.2
Pipe-based Context Switching Test 1318.5 757.9 0.6
Shell scripts (8 concurrent) 4.0 1.0 0.2
=========
SUM of 6 items 7.9
AVERAGE 1.3
--
Arcady Genkin
"... without money one gets nothing in this world, not even a certificate
of eternal blessedness in the other world..." (S. Kierkegaard)
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?87btdvywsn.fsf>