Les Mikesell wrote:
On Tue, 2006-02-14 at 22:17, Mike McCarty wrote:
Its a 2.4gHz Athlon box.
Something wrong with your 58 second machine there Mike :-(
----
take his commentary with a grain of salt and recognize that he is using
FC-2 which I am sure few will argue was better abandoned for FC-3
Are you suggesting that FC4 would load it any faster?
Probably but I think you have a specific problem on that machine.
I suspect the specific problem I have is lack of RAM.
Or too many programs loaded. Adding RAM is easier than sorting
out what you don't really need to run, though.
I've got a dual pII-450 running FC1 that starts OO in about 10
seconds. The first instance of Firefox with google as the home
page takes about 5 seconds, another window about 2. Does
hdparm -T -t show somewhere near 30MB/sec for buffered reads?
# hdparm -T -t /dev/hda
/dev/hda:
Timing buffer-cache reads: 1096 MB in 2.00 seconds = 547.54 MB/sec
Timing buffered disk reads: 66 MB in 3.02 seconds = 21.86 MB/sec
Is that close enough to 30MB/s?
It's less than my ancient PII-450 gets with an almost-as-old
66 mHZ IDE card. A 2.4 gHZ P4 that should be closer to your
specs gets 60MB/s. Do you see something like:
hda: 234441648 sectors (120034 MB) w/8192KiB Cache, CHS=16383/255/63,
UDMA(100)
if you look at 'dmesg | less'? (The UDMA and speed being
the relevant part).
Well, I unloaded everything from memory and did a test.
Nothing running except one "terminal" window, memory usage order.
top - 23:42:46 up 7 days, 5:42, 2 users, load average: 0.21, 0.44, 0.34
Tasks: 73 total, 1 running, 72 sleeping, 0 stopped, 0 zombie
Cpu(s): 1.7% us, 0.3% sy, 0.0% ni, 98.0% id, 0.0% wa, 0.0% hi, 0.0% si
Mem: 248088k total, 176968k used, 71120k free, 9752k buffers
Swap: 524120k total, 116036k used, 408084k free, 66088k cached
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
4159 root 15 0 176m 24m 3512 S 1.3 10.3 965:11.32 X
4536 jmccarty 15 0 34008 11m 4388 S 0.3 4.7 4:26.04 gnome-terminal
4466 jmccarty 15 0 97840 7856 3600 S 0.0 3.2 1:23.81 nautilus
4474 jmccarty 25 10 33040 5968 2348 S 0.0 2.4 2:28.74 rhn-applet-gui
4514 jmccarty 15 0 21776 4704 3132 S 0.0 1.9 3:49.59 wnck-applet
4407 jmccarty 15 0 14640 4356 3336 S 0.0 1.8 7:55.35 metacity
=======================================================================
Loading OO took 24 seconds. So simply low RAM is not the complete
answer. Exiting OO and then restarting it took about 8 seconds,
which corresponds to what others have reported. So it looks like
perhaps a disc bandwidth problem. But I don't know what the cause
would be. At 20 MB/s and achieving, say 20% actual throughput,
22.3% of 256MB is 57MB, which would take at 4 MB/s 14 secs. Hmm,
actually not far off (<2x). I really did do that computation based on
my experience with the disc throughput achieved on other systems
I've used. Hmm. But flushing to virtual should only really double
the time, and 2x14 is not *very* close to 58.
OO loaded, again memory usage order.
top - 23:45:03 up 7 days, 5:44, 2 users, load average: 0.35, 0.41, 0.34
Tasks: 74 total, 1 running, 73 sleeping, 0 stopped, 0 zombie
Cpu(s): 1.0% us, 0.7% sy, 0.0% ni, 98.0% id, 0.0% wa, 0.3% hi, 0.0% si
Mem: 248088k total, 241832k used, 6256k free, 11456k buffers
Swap: 524120k total, 115968k used, 408152k free, 110196k cached
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
3214 jmccarty 15 0 181m 54m 36m S 0.0 22.3 0:07.31 soffice.bin
4159 root 15 0 177m 25m 3640 S 1.0 10.5 965:14.26 X
4536 jmccarty 15 0 34008 11m 4448 S 0.3 4.7 4:26.97 gnome-terminal
4466 jmccarty 16 0 97840 7884 3600 S 0.0 3.2 1:23.90 nautilus
4474 jmccarty 25 10 33040 5988 2348 S 0.0 2.4 2:28.78 rhn-applet-gui
4514 jmccarty 16 0 21776 4712 3136 S 0.0 1.9 3:49.83 wnck-applet
Mike
--
p="p=%c%s%c;main(){printf(p,34,p,34);}";main(){printf(p,34,p,34);}
This message made from 100% recycled bits.
You have found the bank of Larn.
I can explain it for you, but I can't understand it for you.
I speak only for myself, and I am unanimous in that!
