🚀 go-pugleaf

In article <8kVZoB1w164w@coldbox.cojones.com> system@coldbox.cojones.com
(Bryan Lockwood) writes:
# I know there seem to be some technical problems at uunet right now, but I
# wonder if a few sites could post (or email) their normal CPS rates in PEP
# mode for me?
# 
# My problem is that I seem to get pretty good cps on files going TO uunet, but
# files being received FROM uunet are always much slower. Averages:
# 
# ~1300 cps sending files TO uunet
# ~350 cps receiving files FROM uunet
# 
# These rates are for same sessions- by which I mean that in one session, the
# average of all files received will be in the neighborhood of 350 cps, and the
# average of files sent will be in the much higher 1200 range or so. Strange!
# 
# This has been going on since I received muy uunet feed a few months ago.
# 

I have one year's worth of CPS rates for sending and receiving in
PEP mode with UUCP spoofing enabled.  My rates are comparable to
yours, although, once in a blue moon, I do seem to get over 1200
CPS to UUNET.  I wanted to study the data more and figure out if
there is a day of the week or month that is particularly bad to
try to send data to UUNET.  

Right now the data is in "uutraf" format, and I need to run it
through awk or perl to make it more readable.  If you'd like the
uutraf version, I will send it to you.  If you want the prettier
version, you'll have to wait until I am working less than 60
hours per week and have some time.  

			-Jim
-- 
Jim Logan
logan@gimlet.herndon.va.us     ...There's no place like $HOME.
uunet!gimlet!logan

Gee, with basic V.32 (no compression) I get about 750 Long term average
*from* uunet, and 200 long term *to* uunet.  Why the complete reversal?

-- 
---
 Robert Withrow, R.W. Withrow Associates, Swampscott MA 01907 USA
 Tel: +1 617 598 4480, Fax: +1 617 598 4430, Net: witr@rwwa.COM

In article <8kVZoB1w164w@coldbox.cojones.com> system@coldbox.cojones.com (Bryan Lockwood) writes:
>My problem is that I seem to get pretty good cps on files going TO uunet, but
>files being received FROM uunet are always much slower. Averages:
>
>~1300 cps sending files TO uunet
>~350 cps receiving files FROM uunet

Be aware the Telebit modems operating in PEP mode with UUCP spoofing
will buffer data and bias the statistics.  The UUCP "g" spoofing will
buffer many K of data and generate the acknowledments on the local
modem.  This makes transmission look fast.  UUCP thinks that it has sent
the whole file but doesn't know part of it is still buffered in the
local modem.  So the stats it reports look great.  Of course then there
is a big delay waiting for the buffer to empty before the next transfrer
can happen.  Try ignoring the UUCP times and just lock at the log times
between files.

In receive mode the buffering doesn't effect what the receiving host
measures as the thruput.  It is measuring only what it cares about but
the stats seen on the other end could be compleatly different.

So the point is that the 1300 cps is probably only a measure of how fast
your host can talk to your Telebit.  The actual transfer rate to UUNET
could be a lot less.

Regarding cps rates to/from uunet: I'm running ESIX V.4, SAS driver,
HUB6 with 16550AFN's - On long files FROM uunet, my rates are varying
from 950-1100 cps, PEP through a TB+. When uunet is loaded, the rates
drop. From Wed. thru Fri., when traffic is heavier, the rates drop; also,
the rates drop depending on time of day. On a good session, large files
come over at 1100-1175 cps; on a bad one, 600-800 or less. But mostly, I
get over 900.

Your VERY low rates indicate a problem on your end. Are you seeing lots of
"alarms" from your uucico? (If you don't know, invoke uucico for your
uunet sessions with "-x5" to get debug output.) If you are getting lots
of alarms most of the time, especially on long files, the problem is
most likely that your machine isn't fast enough -- your end isn't ACK'ing
packets quickly enough, so uunet goes off to do other things until your
machine sends an "alarm" packet. Then uunet comes back and sends stuff to
you until you miss another ACK, etc.

After I went to V.4 and SAS but before installing 16550's, my throughput
dropped to about 700 cps average on receive -- the characteristic symptom
was lots of alarms on every large file.

Hope this helps.

John Rossmann		(uunet!synsys!john)

Quoted from <1992Aug5.204742.14191@rwwa.COM> by witr@rwwa.COM (Robert Withrow):
> Gee, with basic V.32 (no compression) I get about 750 Long term average
> *from* uunet, and 200 long term *to* uunet.  Why the complete reversal?
> 
> -- 
> ---
>  Robert Withrow, R.W. Withrow Associates, Swampscott MA 01907 USA
>  Tel: +1 617 598 4480, Fax: +1 617 598 4430, Net: witr@rwwa.COM

	I get the same results you do.  I get a 900 Long term average from
UUNET, but get a 200 long term to UUNET.  I am also wondering what it can
be.  Are you connecting via 800 Number or are you using Compuserve lines 
like I am?  I am trying several packet sizes, but think it will not help by
much.  I am connecting through compuserve using a 9600 local line which has
V.42Bis with compression enabled.

	Suggestions??

Carlos
--
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%    __     %  UUCP:  overlord@megalith.miami.fl.us  % I Tried MS-DOS Once %
% __///     %  Fido:  1:135/92                       % But Didn't Inhale.  %
% \\//Amiga %  GEnie: C.Amezaga                      %        --D.Atkins-- %
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

Hello.

FYI, the attached figure is is what I see in terms of receive rate
(char /sec ) recorded by Taylor UUCP between Virginia, U.S.A. and
Tokyo, Japan.  (I use stock SUN UUCP about 1/3 of the time, just in
case there is any performance difference. So far, no siginificant
difference).

Set up is Sun3, TB, SunOS4.0.3, Taylor UUCP 1.03.

Only receiving of files larger than 10000 bytes are recorded. Smaller
files are usually received slower in terms of CP rates due to protocol
overhead..

I have the same observation that sending to UUNET was generally faster
than receiving. But, in the same period (from June up to now), there
wasn't much sending out of big files. So I decided to not to produce
the sending CP rate.  (I attach the gawk script at the end to produce
the output.)

From what I can see from the CP figures below , the variation is
essentially random from cursory observation. JST 19:30 session seems
to offer better output, but not always :-(

I wonder if these variations are caused by
	- just having a very bad connection (telephone, that is),
	- or improper setting of TB? (being looked into), 
	- or there are UUNET CPU/port/modems with different
	  personalities. (I am calling 001,1,703,641,8735).
	- or could modem be left in a strange state with a previous 
          connection???

(I wonder if I should modify Taylor UUCP to drop connection if the
expected througput is less than 500 CP, and call UUNET later. Probably
not a good manner from the viewpoint of UUNET maintenance, though.)

I wonder what other long distance users' experieces are like.

Hope this is entertaining.

Regards,


------------------------------------------------------------
DATE	     Total   CP  A bar graph based on CP 
(Japanese    Volume	(BBB --- very LONG, took me more than > 500 sec!)
Standard     received.  (xxx --- not so good CP rate. :-(
Time)                   (ooo --- relatively good CP rate.)
-------------------------------------------------------------
06-19 19:29  22292   1333 ooooooooooooooooooooooooooo	Very good!
06-19 19:30  24037    735 ***************               
06-19 19:30  12295    814 ooooooooooooooooo
06-19 19:31  11273    845 ooooooooooooooooo
06-19 19:32  92279    937 ooooooooooooooooooo
06-20 03:29  39837    893 oooooooooooooooooo
06-20 03:29  16360    833 ooooooooooooooooo
06-20 03:30  13784    830 ooooooooooooooooo
06-20 03:30  21237    787 ****************
06-20 03:31  11246    799 ****************
06-20 03:31  19107    854 oooooooooooooooooo
06-20 19:28  11556   1082 oooooooooooooooooooooo
06-20 19:32 170360    930 ooooooooooooooooooo
06-20 19:33  93187    930 ooooooooooooooooooo
06-21 03:29  17644    385 xxxxxxxx
06-21 03:30  17669    375 xxxxxxxx
06-21 19:28  18706   1262 oooooooooooooooooooooooooo		<---- ditto.
06-21 19:30  65567    787 ****************
06-23 22:40 152223    394 xxxxxxxx
06-24 03:29  17947    824 ooooooooooooooooo
06-24 19:30  32246   1350 oooooooooooooooooooooooooooo		<---- ditto.
06-24 19:30  13161    829 ooooooooooooooooo
06-24 19:33 125197    849 ooooooooooooooooo
06-25 20:01 139703    957 oooooooooooooooooooo
06-25 23:11  11600    300 xxxxxxx
06-26 11:29  12832    823 ooooooooooooooooo
06-26 11:30  14652    825 ooooooooooooooooo
06-26 11:31  11256    579 ************
06-26 21:49 250923    343 BBBBBBB	<--- DISASTER strikes! 
06-26 21:49  16239    320 xxxxxxx
06-27 11:32  12970    366 xxxxxxxx
06-27 11:34  26204    391 xxxxxxxx
06-27 11:35  14699    381 xxxxxxxx
06-28 03:33 156727    932 ooooooooooooooooooo
06-28 03:33  18658    834 ooooooooooooooooo
06-28 03:34  12391    794 ****************
06-28 11:28  13273    813 ooooooooooooooooo
06-28 11:29  36602    864 oooooooooooooooooo
06-28 11:30  49022    873 oooooooooooooooooo
06-28 19:29  12671    362 xxxxxxxx
06-28 19:30  18227    386 xxxxxxxx
06-29 11:30  19555    422 *********
06-29 11:31  33554    420 *********
06-30 03:29  13467    831 ooooooooooooooooo
06-30 03:30  33465    892 oooooooooooooooooo
06-30 03:31  10578    781 ****************
06-30 03:31  17914    853 oooooooooooooooooo
06-30 03:32  14403    780 ****************
07-01 03:28  11571    792 ****************
07-01 03:29  12123    803 ooooooooooooooooo
07-01 03:29  18607    848 ooooooooooooooooo
07-01 03:30  25787    733 ***************
07-01 03:30  18988    853 oooooooooooooooooo
07-02 11:47 199819    275 BBBBBB	<--- ditto.
07-02 11:48  16111    316 xxxxxxx
07-02 11:49  11776    311 xxxxxxx
07-02 11:51  16962    316 xxxxxxx
07-02 11:52  14960    314 xxxxxxx
07-03 03:30  34656   1430 ooooooooooooooooooooooooooooo	<--- good!
07-03 03:35 297552    956 oooooooooooooooooooo
07-03 03:36  12820    806 ooooooooooooooooo
07-03 03:36  13619    832 ooooooooooooooooo
07-03 03:37  29982    879 oooooooooooooooooo
07-03 03:37  13801    828 ooooooooooooooooo
07-03 03:38  22019    869 oooooooooooooooooo
07-04 03:32  19153    307 xxxxxxx
07-05 03:30  35621    759 ****************
07-05 03:36 147939    413 *********
07-05 03:41 107359    411 *********
07-05 03:44  13576    381 xxxxxxxx
07-05 11:30  11725    370 xxxxxxxx
07-05 19:28  16334    804 ooooooooooooooooo
07-05 19:30  18729    809 ooooooooooooooooo
07-05 19:30  14207    791 ****************
07-06 19:32  14788    384 xxxxxxxx	<--- makes me feel gloomy....
07-06 19:33  12316    371 xxxxxxxx      <--- to look at all these...
07-06 19:34  10255    379 xxxxxxxx
07-06 19:35  16853    386 xxxxxxxx
07-06 19:37  18159    383 xxxxxxxx
07-06 19:38  32925    390 xxxxxxxx
07-07 19:39 190099    395 xxxxxxxx
07-07 19:39  10396    361 xxxxxxxx
07-07 19:41  12743    365 xxxxxxxx
07-07 19:42  14031    370 xxxxxxxx
07-08 03:29  24263    872 oooooooooooooooooo
07-08 03:30  31178    872 oooooooooooooooooo
07-09 03:29  17214    818 ooooooooooooooooo
07-09 03:29  11051    717 ***************
07-09 03:29  15597    868 oooooooooooooooooo
07-09 03:30  14167    798 ****************
07-10 03:38 219115    405 BBBBBBBBB	<--- AGAH...
07-10 03:38  12464    363 xxxxxxxx
07-10 03:40  19078    377 xxxxxxxx
07-10 03:41  21088    381 xxxxxxxx
07-10 03:41  12947    374 xxxxxxxx
07-10 03:42  13121    375 xxxxxxxx
07-10 03:43  25007    388 xxxxxxxx
07-10 19:34 269137    897 oooooooooooooooooo
07-11 03:31  31270    567 ************
07-11 03:33  23024    309 xxxxxxx
07-11 03:34  10490    290 xxxxxx
07-11 19:29  33189   1356 oooooooooooooooooooooooooooo <--- Phew!
07-11 19:31 110691    940 ooooooooooooooooooo
07-12 03:28  21647    878 oooooooooooooooooo
07-12 11:28  14857    795 ****************
07-12 11:29  18262    851 oooooooooooooooooo
07-12 11:29  15614    697 **************
07-12 19:29  17613    844 ooooooooooooooooo
07-13 03:29  16535    394 xxxxxxxx
07-13 03:30  14375    383 xxxxxxxx
07-13 03:30  10628    368 xxxxxxxx
07-13 11:29  13329    791 ****************
07-13 19:30  26952    738 ***************
07-14 11:32  11101    377 xxxxxxxx
07-14 19:31 112363    929 ooooooooooooooooooo
07-15 11:31  23285    817 ooooooooooooooooo
07-15 11:32  71391    880 oooooooooooooooooo
07-15 11:33  15569    810 ooooooooooooooooo
07-15 11:33  21601    828 ooooooooooooooooo
07-15 11:34  11891    754 ****************
07-15 11:34  15539    718 ***************
07-15 11:35  14088    643 *************
07-15 19:39 174637    335 BBBBBBB	<--- These bad CP cosume by budget!
07-16 03:29  24727    763 ****************
07-16 03:29  11907    779 ****************
07-16 03:30  16486    696 **************
07-16 03:31  17627    796 ****************
07-17 03:30  90665    928 ooooooooooooooooooo
07-17 03:31  34746    892 oooooooooooooooooo
07-17 03:32  51919    908 ooooooooooooooooooo
07-17 03:33  38580    888 oooooooooooooooooo
07-17 03:33  19249    858 oooooooooooooooooo
07-17 03:34  19709    812 ooooooooooooooooo
07-18 03:32  58399    313 xxxxxxx
07-18 03:37  78045    325 xxxxxxx
07-18 03:38  17527    307 xxxxxxx
07-18 03:42  72073    325 xxxxxxx
07-19 03:36 185042    919 ooooooooooooooooooo
07-19 03:36  10788    688 **************
07-19 03:37  16391    837 ooooooooooooooooo
07-19 19:37 181856    648 *************
07-19 19:38  18875    568 ************
07-19 19:40  36357    739 ***************
07-20 11:30  14766    813 ooooooooooooooooo
07-20 11:31  30233    830 ooooooooooooooooo
07-20 19:36 241895    695 **************
07-21 03:32  15009    364 xxxxxxxx
07-21 03:33  10191    348 xxxxxxx
07-21 03:35  21706    369 xxxxxxxx
07-21 19:34  99085    677 **************
07-22 03:43 890302    979 BBBBBBBBBBBBBBBBBBBB	<--- STUPID user tries
07-22 03:58 890302    980 BBBBBBBBBBBBBBBBBBBB  <--- copying a big patch,
07-22 03:59  10633   1014 ooooooooooooooooooooo   TWICE. But CP WAS GOOD.
07-22 04:01  29640    896 oooooooooooooooooo
07-23 11:34  11620    306 xxxxxxx
07-24 03:35 120583    392 xxxxxxxx
07-24 03:36  12202    361 xxxxxxxx
07-24 03:37  26258    389 xxxxxxxx
07-24 03:41  83481    389 xxxxxxxx
07-24 03:43  37564    382 xxxxxxxx
07-24 03:43  11048    371 xxxxxxxx
07-24 03:44  10300    365 xxxxxxxx
07-24 03:46  39645    385 xxxxxxxx
07-24 03:47  17557    375 xxxxxxxx
07-25 11:52 333989    334 BBBBBBB	<--- hmm....
07-25 11:54  23076    314 xxxxxxx
07-25 11:56  11044    305 xxxxxxx
07-25 11:57  11642    308 xxxxxxx
07-25 11:58  10978    309 xxxxxxx
07-25 11:59  13919    303 xxxxxxx
07-25 19:30  85665    895 oooooooooooooooooo
07-26 03:33  97449    392 xxxxxxxx
07-26 03:33  15761    374 xxxxxxxx
07-26 11:28  14860    833 ooooooooooooooooo
07-27 19:32  37694    889 oooooooooooooooooo
07-27 19:33  11234    817 ooooooooooooooooo
07-27 19:35  22996    881 oooooooooooooooooo
07-27 19:36  17804    820 ooooooooooooooooo
07-27 19:38  77473    843 ooooooooooooooooo
07-28 03:31  19786    868 oooooooooooooooooo		
07-29 03:38 165960    404 *********			
07-29 03:39  21027    382 xxxxxxxx
07-29 03:40  17223    382 xxxxxxxx
07-29 03:40  10648    365 xxxxxxxx
07-29 03:41  15431    384 xxxxxxxx
07-29 03:42  20637    388 xxxxxxxx
07-30 03:33 208361    952 oooooooooooooooooooo
07-30 03:34  41211    899 oooooooooooooooooo
07-30 03:34  11732    805 ooooooooooooooooo
07-30 03:35  24984    847 ooooooooooooooooo
07-30 03:36  20406    796 ****************
07-30 11:31  19459    309 xxxxxxx			
07-31 03:51 242245    218 BBBBB	<--- I feel like crying here.:-(	
07-31 03:52  14080    213 xxxxx <--- CP rate < 300 is BAAAAAAAAD!
07-31 03:53  11062    213 xxxxx (Linda Ronstadt, or Michael Jackson tune
07-31 03:55  16962    214 xxxxx  comes to my mind. Bad...)
07-31 03:56  13973    211 xxxxx
07-31 03:58  16964    214 xxxxx
08-01 03:28  10246    781 ****************
08-01 19:46 191455    216 BBBBB		<--- ditto.
08-02 03:28  11295    364 xxxxxxxx
08-02 03:29  12657    356 xxxxxxxx
08-02 19:33  15224    213 xxxxx
08-02 19:34  12613    207 xxxxx
08-03 19:30 114075    946 ooooooooooooooooooo
08-04 03:29  10190    774 ****************
08-04 03:29  10214    748 ***************
08-04 03:30  11904    785 ****************
08-04 03:30  12176    791 ****************
08-05 03:35 133556    405 *********
08-05 03:37  10406    360 xxxxxxxx
08-05 03:38  15289    375 xxxxxxxx
08-06 03:31 120905    928 ooooooooooooooooooo	<--- today, so far so good.:-)
08-06 03:32  19513    824 ooooooooooooooooo	  
08-06 03:32  16577    856 oooooooooooooooooo
08-06 03:33  14769    824 ooooooooooooooooo
08-06 03:33  14074    802 ooooooooooooooooo


(cf. The following is the awk script I used to make this output.
#    The field number probably ought to be fiddled for your Taylor
#    output format.
#
#	analyze.awk
#
#    analyze STATFILE
#    STATFILE is the statistical output from Taylor UUCP.
#    Actually you MUST 
#             egrep uunet ORIGINAL_STATFILE > STATFILE
#    to produce the suitable UUNET only output file.
#
#    gawk -f thisfile STATFILE will produce receive rate fig.
#    gawk -f thisfile flag=1 STATFILE will produce send rate fig.
#

BEGIN {
	stars="*************************************************"
	ostars="oooooooooooooooooooooooooooooooooooooooooooooooo"
	xstars="xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx"
	Xstars="BBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBB"

	}
flag == 0 && $5=="received" && ($6 > 10000) {
	rate = substr($11,2)
	if($9 > 500)
		marks = Xstars
	else if(rate < 400)
		marks = xstars
	else if (rate < 800)
		marks = stars
	else 
		marks = ostars
	date = substr($3,7) " " substr($4,1, 5)
	volume = $6
	
	printf ("%s %6d %6d %s\n", date, volume, rate, substr(marks,1, (rate/50)+1));
}
flag > 0 && $5=="sent" && ($6 > 10000) {
	rate = substr($11,2)
	if($9 > 500)
		marks = Xstars
	else if(rate < 400)
		marks = xstars
	else if (rate < 800)
		marks = stars
	else 
		marks = ostars
	date = substr($3,7) " " substr($4,1, 5)
	volume = $6
	
	printf ("%s %6d %6d %s\n", date, volume, rate, substr(marks,1, (rate/50)+1));
}


END {
	dummy =0
}
--
------------------------------------------------------------------
    Chiaki Ishikawa                       Phone:+81-3-5702-0351
  Personal Media Corp.                      FAX:+81-3-5702-0359
 MY Bldg, 1-7-7 Hiratsuka,   UUNET: ... uunet!pmcgw!ishikawa, or
 Shinagawa-ku, Tokyo 142                ishikawa%pmcgw@uunet.uu.net
        JAPAN                JUNET:     ishikawa@personal-media.co.jp
------------------------------------------------------------------

>In article <8kVZoB1w164w@coldbox.cojones.com> system@coldbox.cojones.com
>(Bryan Lockwood) writes:
># I know there seem to be some technical problems at uunet right now, but I
># wonder if a few sites could post (or email) their normal CPS rates in PEP
># mode for me?
># 
># My problem is that I seem to get pretty good cps on files going TO uunet, but
># files being received FROM uunet are always much slower. Averages:
># 
># ~1300 cps sending files TO uunet
># ~350 cps receiving files FROM uunet
># 
># These rates are for same sessions- by which I mean that in one session, the
># average of all files received will be in the neighborhood of 350 cps, and the
># average of files sent will be in the much higher 1200 range or so. Strange!
># 
># This has been going on since I received muy uunet feed a few months ago.
># 
 
All of these sites connect to us with PEP mode.  aee is a local
call (wish more of these were <grin>) over standard CenTel lines.
All other calls egress or enter via our digital lines to the
long distance carrier we have (Telecom*USA/MCI).  

Baud rate on my side is locked at 19200, using T2500SA rom GE7.03.
aee is a T2500SA GF7.03, gator is a World Blazer, ksmith is a
Telebit of some kind, tssi is a World Blazer, and uunet is 
whatever-they-use. :)  UUCP Spoofing is enabled in non-vol RAM.

Hope this helps.  (This is traffic over the last 8 hours, and
agree with figures that get mailed nightly to uucp before the
nightly cleanup script runs.)

:set ts=8

aee!summary
				Total		Out		In
	Characters transferred: 1841514     	1840689     	825         
	  Connect time (secs.): 1778.020 	1772.320 	5.700    
	 Average transfer rate: 1035.7
 
gator!summary
				Total		Out		In
	Characters transferred: 1689653     	1538161     	151492      
	  Connect time (secs.): 1581.290 	1333.920 	247.370  
	 Average transfer rate: 1068.5
 
ksmith!summary
				Total		Out		In
	Characters transferred: 8819573     	8819573     	0           
	  Connect time (secs.): 7214.070 	7214.070 	0.000    
	 Average transfer rate: 1222.6
 
tssi!summary
				Total		Out		In
	Characters transferred: 5981479     	713117      	5268362     
	  Connect time (secs.): 4753.050 	582.170  	4170.880 
	 Average transfer rate: 1258.5
 
uunet!summary
				Total		Out		In
	Characters transferred: 2663963     	304817      	2359146     
	  Connect time (secs.): 2560.930 	263.470  	2297.460 
	 Average transfer rate: 1040.2

-- 
Andrew Finkenstadt       | Vista-Chrome, Inc.   |           andy@GEnie
GEnie Unix Sysop/Manager | The Printing House   |    NIC Handle: AF136
+1 904 222 2639 home     | 1600 Capital Cir SW  |   ...!uunet!rde!andy
+1 904 575 0189 work     | Tallahassee FL 32310 | andy@vistachrome.com