LabVIEW
find network ip addresses
CB wrote:
> Hi All,
> I would like to generate a list of the ip addresses on a network to
> allow the user to choose 1 or many machines to pass data to, from 1
> of the network machines. Data flow will be 115200 bps.
> Thanks in advance, Chris
Why 115200 bps? Sounds like a serial port to me but then I don't
understand what you are talking about with network access. Anyhow the
first idea is tricky. There is no simple way to get this information.
Basically TCP/IP is all about knowing both the server and client BEFORE
you make a connection.
The only thing I would see is to create on one system a service which
listens for UDP messages from other applications. All the potential
clients which your user can connect to broadcast UDP announcements about
their
presence when starting up and in regular intervals. The UDP
service collects these messages and maintains a list of those clients.
Your user machine can then query this list, either through an well
configured TCP/IP connection (safe) or also through an UDP query
(unreliable but you don't need to configure the UDP service machines
address) and presents this list to the user.
There are some other issues with this. The list needs to be regularly
cleaned somehow to purge computers not anymore alive. It is additional
overhead to write these services etc.
Another way is if your machine on which you may need to select other
computers stays always the same: Let the clients on startup contact your
well known machine over TCP/IP and have them send a message that they
are alive. Maintain that list in your server application and allow user
machines to connect it to retrieve this cleint list as well.
Rolf Kalbermatter
> Hi All,
> I would like to generate a list of the ip addresses on a network to
> allow the user to choose 1 or many machines to pass data to, from 1
> of the network machines. Data flow will be 115200 bps.
> Thanks in advance, Chris
Why 115200 bps? Sounds like a serial port to me but then I don't
understand what you are talking about with network access. Anyhow the
first idea is tricky. There is no simple way to get this information.
Basically TCP/IP is all about knowing both the server and client BEFORE
you make a connection.
The only thing I would see is to create on one system a service which
listens for UDP messages from other applications. All the potential
clients which your user can connect to broadcast UDP announcements about
their
presence when starting up and in regular intervals. The UDP
service collects these messages and maintains a list of those clients.
Your user machine can then query this list, either through an well
configured TCP/IP connection (safe) or also through an UDP query
(unreliable but you don't need to configure the UDP service machines
address) and presents this list to the user.
There are some other issues with this. The list needs to be regularly
cleaned somehow to purge computers not anymore alive. It is additional
overhead to write these services etc.
Another way is if your machine on which you may need to select other
computers stays always the same: Let the clients on startup contact your
well known machine over TCP/IP and have them send a message that they
are alive. Maintain that list in your server application and allow user
machines to connect it to retrieve this cleint list as well.
Rolf Kalbermatter
