Re: Sockets, port and loop-back ?
From: P.T. Breuer (ptb_at_oboe.it.uc3m.es)
Date: 01/06/04
- Next message: jack: "Re: sshing into computers behind a router"
- Previous message: Fred Stuurman: "Re: fetchmail multidrop"
- In reply to: not_at_top-post: "Re: Sockets, port and loop-back ?"
- Next in thread: jack: "Re: Sockets, port and loop-back ?"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: Tue, 06 Jan 2004 10:20:17 GMT
not@top-post wrote:
> not@top-post wrote:
>[attribs lost]
> > > at the IP of <someID>, and http://
> > > will address the <not default number> port at the IP of <someID> ?
> Then could the scarcity of IPs be temporarily alleviated by
Scarcity?
> doubling-up on IP's by using some of the unused ports in a
> non-standard way ? Of course this would require addition of
> knowledge/intelligence, whereas the trend is towards dumbing
> down towards more convenience.
> One wouls have eg:
> http://
> http://
You could do that, but why bother? You can have as many "virtual" http
servers as you like on the SAME, standard port, simply by pointing
different names at the the same IP address. The server will distinguish
which personality it should take according t the name by which it is
addressed.
You'll have to identify a more specific problem that you are thinking
of.
> Is there a problem if IPs have multiple names for 'reverse
> DNS-ing' ?
No.
> ]AFAIK, the operating system itself requires a loopback network, the
> ]address of which is traditionally 127.0.0.1. It serves no other purpose
> ]than to allow software to interact with the os via ports on that
> ]network. Or do I not understand this matter?
> In the case that my in & out to the network is via my serial port
> [to my (duplex) modem], I don't have the serial port 'looped back'.
Lucky you, or it wouldn't go anywhere!
> So where is the loopback implemented ?
What loopback? The loopback IP address range is a "virtual" address
range handled by your kernel. Packets directed there are handled by
your own kernel and never leave your machine.
> I'm guessing that the
> send and receive buffers for the serial port are 'looped back',
Why guess? If they were, you could claim your money back!
> ie. just connected - taking the serial-port out of the loop.
Uh - I fail to decipher exactly what you could be thinking of, but
whatever it is, it is nowhere near true. Are you asking how you turn a
serial port into a loopback connection for testing purposes? You
connect its control lines round to each other. See a serial howto, or
some such.
Peter
- Next message: jack: "Re: sshing into computers behind a router"
- Previous message: Fred Stuurman: "Re: fetchmail multidrop"
- In reply to: not_at_top-post: "Re: Sockets, port and loop-back ?"
- Next in thread: jack: "Re: Sockets, port and loop-back ?"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|
