[aprssig] CQ Field Day via CQSRVR
bruninga at usna.edu
Sun Oct 21 12:09:20 CDT 2007
> Actually no "melt-down" will occur. If all
> 20,000 APRS users (count for over the past
> 30 days) tried to do a CQ FIELDAY, CQSRVR,
> APRS-IS, and the respective local APRS RF
> networks could handle it.
I don't think so. Lets take just 10% of that number, say 2000 field day sites. Worst case scenario, is they CQ every 30 minutes. CQSRVR then sends all 2000 of them to all 2000 other user. So each user in each RF area then gets 2000 packets every 30 minutes in his own local RF area. In my mind that is melt down. Especially if more than one FD site is in the same RF area and Igates are notorius for duplicating packets.
In the central Maryland RF area, I would guess there are 15 active FD sites. Probably served by 5 or so IGates. So 15 sites, each getting 2000 packets duplicated maybe by at least 2 IGates on average would be a total of 60,000 extra packets every 30 minutes.
A melt down of the 1200 baud channel indeed.
So before Field Day, some kind of automatic threshold limiting process will be needed, or we will be shooting ourselves in the foot with grid lock and nothing getting through.
Using the bin counting/limiting approach would limit and spread out the load and prevent groid lock in the RF domains.
More information about the aprssig