I do understand your problems but I see no way in reasonably filtering
or throttling the DX spot feeds. Your choice would either be in
disconnecting the internet during prime time or really struggling for
an upgrade of your link capabilities. Isn't the current situation you are
facing a good reason to show that 1200 baud links are not enough?
WU4G posted an interesting message in the Packet Cluster sysop's
mailing list which I want to forward to you in full length (sorry
for those who are on both lists.)
> After seeing the comments about spots being so far behind; watching our
> spots do the same thing here due to the volume; and hearing the same comment
> made about ways to cut out spots from some of the more distant nodes, it
> struck me that for such a "technologically advanced" group (U.S. hams
> especially here), we sure are behind the stick on this one. We work to
> build our links, make megaclusters and how do we handle the traffic? At a
> stinking 1200 baud! Regular computer modems can now handle 56K and we're
> still in the dark ages. No wonder the kids laugh at us today when we talk
> about the wonders of packet radio. Many of the European clusters have long
> since left 9600 baud in the dirt, what is our problem ? Lack on interest
> in furthering the packet technology? No desire to spend the money? Our
> manufacturers marketed 2400 and 9600 baud TNCs but sales for those products
> are in the toilet. They have no incentive to develop anything faster and I
> can't blame them for that. We shouldn't be trying to figure how to cut
> traffic so what little info does get through is more timely, we should be
> striving to handle more and more traffic efficiently. One day we are going
> to receive a rude awakening when we volunteer our services in the interest
> of public service only to be laughed at when we demonstrate our measly 1200
> baud packet links and they pull out their high speed digital satellite link.
>
> This wasn't a soapbox to anyone specific, but to all of us, myself included.
> Thanks for the new software announcement. I think it's great. It shows we
> still have the desire to further the technology. Now all we need is the
> desire to install it in something capable of handling that technology.
>
> 73 and hope exchange spots in real time !
> Ron WU4G
Here are some statistics from DB0MDX which run well throughout last
weekend with 50 or more users and 8 links.
Number of DX Spots over the last 7 days:
Nov 24 00:00:00 1998 GMT - Nov 24 23:59:59 1998 GMT : 619
Nov 25 00:00:00 1998 GMT - Nov 25 23:59:59 1998 GMT : 753
Nov 26 00:00:00 1998 GMT - Nov 26 23:59:59 1998 GMT : 765
Nov 27 00:00:00 1998 GMT - Nov 27 23:59:59 1998 GMT : 1177
Nov 28 00:00:00 1998 GMT - Nov 28 23:59:59 1998 GMT : 3546 <--- CQWW-CW
Nov 29 00:00:00 1998 GMT - Nov 29 23:59:59 1998 GMT : 4394 <--- CQWW-CW
Nov 30 00:00:00 1998 GMT - Nov 30 23:59:59 1998 GMT : 859
Dec 01 00:00:00 1998 GMT - Dec 01 23:59:59 1998 GMT : 768
This extended statistic function will be available in the db_maint tool
in release 4.03.
Calculating 3546 spots (let's say 80 bytes in length) times 50 users
gives ~14.2 MB of data. Dividing by 60*60*24 (seconds) gives a
rate of 164 Bytes/second which equals 1313 bps. There you have no
AX.25 protocol overhead, no retries, no additional information (like PC19
traffic). I'd estimate that on a single 1200 baud link one cannot have
more than 10 connections simultaneuosly with that huge amount of
traffic.
So you better go for upgrading your links, Ron.
73 Ben, DL6RAI
--
[] clx - PacketCluster Software for Linux
[] c/o Bernhard ("Ben") Buettner, DL6RAI
[] Internet: terhi.victor@logonet.com
[] Packet: DL6RAI @ DB0AAB.#BAY.DEU.EU