[Dxspider-support] Skimmers

[Brendan Minish]

On Sun, 2011-03-20 at 10:41 +0000, gu6efb wrote:
> Hi Group
> 
> What are sysops views on the rise of all theses skinners such as W3LPL
> and UT7MA just to name a few.


It's My view that the Skimmers are extremely useful BUT have no place
feeding spots to the conventional network, under ANY circumstances,
ever.

There is a large network of skimmers. it's called the reverse beacon
network (RBN), My skimmer reports to it along with about 50 others.
http://www.reversebeacon.net/

The system does have telnet ports, one of which runs dxpsider however
the level of traffic would I suspect, bring the conventional cluster
knees.
During CW contests the system handles a truly vast amount of spots. 
Yesterday for example during the RDX contest the RBN spider node reports
that there were 452,529 spots in 1 24 Hour period, My conventional
cluster node handled 17,246 spots   

Both W3LPL and UT7MA are active in the Russian DX contest as
multi-operator stations and seem to be spotting all QSO's from their
Search & Pounce (multiplier)  station(s), This IS useful to other
contesters and a feature that many contest logging programs support. I
have used this on occasion myself but with my S&p run rates it's
unlikley to upset anyone ;-)   

I am not seeing skimmer spots from either station on the regular network
at present, Just S&P spots.
W3LPL is submitting spots to RBN currently at the rate of several per
minute, as are the other connected skimmers  

By the way it's quite a testament to DXspider's abilities and Dirk's
coding that DXSpider can handle 90+ connected users at the very high
spot rates that RBN sees. 
However I suspect that there is room for further performance
optimisations for these traffic rates. 

> I have had a number of complaints about these skinner stations from my
> users.
> 
> I have put the calls of skimmer stations in the badspotter list which
> is helping.

the problem with this is that both stations generate a lot of legitimate
spots too, I suspect it's really a user issue and it might be better
solved by asking stations sending skimmer spots to the regular network
not to do it 

> I know that CC Cluster has a NO Skimmer command is this something that
> we now need in Dx Spider?

VE7CC IS collecting skimmer spots from some skimmers directly, He did
ask me for a connection to my skimmer but at present I don't allow
inbound connections to it, I simply push a connection out to the RBN.
I also think the needs of the DX community are better served by a single
system or network handling skimmer spots rather than disparate systems
each carrying only a portion of the shared skimmers.

It would be nice to see some features in Dxbase to allow easy peering
with nodes supporting skimmer spots so that there is no possibility of
these spots being leaked back to the regular cluster network.
perhaps a variation of the set/node command that allows a peer to be
declared a skimmer nod and set up the filtering rules appropriately 

there would also need to be user commands to give the users control of
which types of spots they see.

I would love to run a node that carried Skimmer spots in addition to
regular spots but only propagate local spots back to the regular cluster
network.
Right now I mix feeds from RBN and my spider node on my local computer
with WintelnetX then feed this to my logging programs 
http://www.k1ttt.net/software.html#wintelnetx
but it's a bit of a hack 

-- 
73
Brendan EI6IZ