| |
N1URO > PACKET 11.01.19 07:58l 51 Lines 2253 Bytes #999 (0) @ WW
BID : 15611_N1URO
Read: GUEST
Subj: Re: PE1ERR > Number of nodes
Path: IW8PGT<IZ3LSV<F1OYP<ON0AR<GB7CIP<N1URO
Sent: 190111/0544Z @:N1URO.#CCT.CT.USA.NOAM #:15611 [Unionville] $:15611_N1URO
From: N1URO@N1URO.#CCT.CT.USA.NOAM
To : PACKET@WW
Red;
> I think you're just deliberatly misunderstanding me in order to take some
> virtual high ground
What purpose would that serve? None. I think you're reading too deeply into
my words.
> Please, explain if you believe it is lack of knowledge, I'm asking for your
> help in understanding.
> Go on... I dare you
If you meant this you wouldn't pose this as a dare, however to give you a
little bit of an answer - when you lock in paths and you lose the path to
a neighbor node whether it be via RF or an internet based path (say your
ISP loses it's BGP routing to that neighbor for whatever reason) and then
another neighbor can still give you a path to that node you yourself lost,
you still risk answering that neighbor's frames through the locked route.
Just as if you had an RF neighbor say on 440Mhz and you locked in a path
on that frequency to them and they lost their 440Mhz radio and/or TNC for
whatever reason but they also had a 220Mhz set up as a backup to you, you
would force any response frames incoming to you on 220 -> 440 because you've
locked in the route rather than allow it to be dynamic in nature.
Here on EastNet we use pc/FlexNet as our ax.25 routers on our RF network
and we route NetRom, IPv4, and IPv6 through it fine and because of how
dynamic the routing is and with the redundant paths we try to maintain
if one path is lost FlexNet will automatically reroute the frames as they
can be routed with the best efficiency as possible.
I've even witnessed 2 xNOS systems exchanging an SMTP type message and in
mid stream the primary path failed however the secondary path automatically
engaged and the initial IP socket that was created never dropped! It was
really very fascinating to watch in a sniff!
Granted NetRom's dynamic routing isn't as quick as FlexNet's is but when you
lock in paths/routes then that's when you in a sense begin to disable what
it can do for you. Another benefit of flexnet is that it's pure ax25 thus
taking full advantage of the 256 byte MTU where using NetRom takes away 20
bytes from the data segment of the frames for it's protocol overhead... thus
making Flex appear faster.
73 de N1URO
---
SendBBS v1.1 by N1URO for LinFBB
Read previous mail | Read next mail
| |
