[Dxspider-support] More about routing.

Rene Olsen rene_olsen at post3.tele.dk
Sat Jan 18 09:22:45 GMT 2003


Hi.

Thanks to Dirk for replying promptly to my last mail. It sounds like work is being done to 
solve this.

In the mean time lets try to clean up the routing. If you link to a clx node with more than 
one active link then don't trust the info that comes from there.

If you have more than one link in your spider, make sure that you don't accept the same 
node info from more than one link. Lets say that you have two or more links to some 
country, then you need to make sure that you only trust the node/user info from that 
country from one of these nodes.

Spider seems to be able to handle routing loops, but how do we expect it to keep track 
of what is correct and what is not ?

Lets take an exanple:

OZ1XXX is linked to ON0XXX with an isolated link. This means that the local users are 
known and transfered on that link.

OZ1XXX has full protocol link with OZ2XXX and OZ2XXX has full protocol link to 
EA5XXX.

What happens now is that at some point the entire OZ node/user list will arrive at 
ON0XXX. Now, what do we expect spider to do. It now has OZ1XXX routing from two 
sources. It has it directly linked with an isolated link, and now it gets the same info from 
somewhere else. How is ON0XXX to decide which info can be trusted ?

In this case it should probably trust the info from the direct link, and then accept the rest 
of the OZ info.

But what if OZ1XXX was not linked directly. What if ON0XXX got the complete OZ 
routing from ON1XXX and ON2XXX. How does it then decide what can be trusted ?

I think we really need to clean up our links, and limit the routing protocol to make sure 
that we have control of what we accept from who. That is the only way to make sure 
that we don't get the same routing info from more than one source.

The alternative is to continue as we do now, and live with the fact that the node/user list 
will never be acurate.

Maybe Dirk can try to shed some light on how spider tries to handle the situation of 
getting the same routing info from several sources atm, and how spider chooses what to 
use and not to use.

Vy 73 de Rene / OZ1LQH






More information about the Dxspider-support mailing list