Great Circle Associates List-Managers
(August 1996)
 

Indexed By Date: [Previous] [Next] Indexed By Thread: [Previous] [Next]

Subject: Re: Reply from Compuserve
From: Dave Bigham <dbigham @ cantec . com>
Date: Tue, 6 Aug 1996 17:28:34 -0700 (PDT)
To: list-managers @ GreatCircle . COM

At 09:06 PM 8/6/96 +0200, you wrote:
>On Tue, 6 Aug 1996 19:06:37 +0200 Kjetil Torgrim Homme
><kjetilho @
 ifi .
 uio .
 no> said:
>>Here is the  kind of statistics you can expect with
>previous generation equipment:
>
>Total: 166,690 deliveries
>
>+--------------------------+--------------------------+------------+
>|  Average delivery time   |     Max delivery time    | For...     |
>+--------------------------+--------------------------+------------+
>|      3 sec       (00:03) |      8 sec       (00:08) | First 50%  |
>|      5 sec       (00:05) |     13 sec       (00:13) | First 60%  |
>|      6 sec       (00:06) |     21 sec       (00:21) | First 70%  |
>|      8 sec       (00:08) |     26 sec       (00:26) | First 80%  |
>|     11 sec       (00:11) |     35 sec       (00:35) | First 90%  |
>|     12 sec       (00:12) |     55 sec       (00:55) | First 95%  |
>|     13 sec       (00:13) |     72 sec       (01:12) | First 96%  |
>|     14 sec       (00:14) |    130 sec       (02:10) | First 97%  |
>|     17 sec       (00:17) |    785 sec       (13:05) | First 98%  |
>|     27 sec       (00:27) |   1596 sec       (26:36) | First 99%  |
>|    216 sec       (03:36) | 334125 sec (3d 20:48:45) | All (100%) |
>+--------------------------+--------------------------+------------+
>
>As  you can  see, half  of the  messages were  delivered in  less than  8
>seconds.  The  average delivery  time,  setting  aside  the usual  1%  of
>addresses for brain-dead  mail systems that support a grand  total of one
>parallel  connection  or are  up  2  hours a  day,  is  27 seconds.  This
>turbo-charged super-server  is a 133MHz  Pentium with 64M and  one drive.
>This machine  also runs  a web  server from which  people can  browse the
>archives and make searches. There are 123 lists ranging from a handful of
>subscribers to 1,400 or so. It's  far from being the largest mailing list
>site in  the world,  but it  can probably  handle the  needs of  99.x% of
>existing mailing  list sites, and it's  really not much hardware  at all.
>It's only a matter of time until most lists run on that kind of system. I
>give it a year, perhaps two.
>
>  Eric
>

Eric,

I think it is important to note that the above statistics are individual
statistics for all of the 166,000+ messages sent from all of the lists. They
are not from one list at one time. So, for example, *each one* of the
fastest 50% took 8 seconds to send - each. Thats more that 664,000
message-seconds, if that makes any sense. It's not a case of 80,000+
messages being sent in 8 seconds or less.

I'm probably stupid, but that's what I read the first time through it.

Dave Bigham


Indexed By Date Previous: Re: Reply from Compuserve
From: Brock Rozen <brozen @ netvoyage . net>
Next: header problem
From: "Steven K. Buehler" <skb @ sbcomp . com>
Indexed By Thread Previous: Re: Reply from Compuserve
From: "Henry W. Miller" <henrym @ sacto . mp . usbr . gov>
Next: Re: Reply from Compuserve
From: jms @ tardis . Tymnet . COM (Joe Smith)

Google
 
Search Internet Search www.greatcircle.com