1

Topic: Nware NioNode device discovery

Hi all,

When does NWare look on the Network for nions/Conman to connect to?

It seems to me that this only happens at startup of Nware. Is there a way to force this discovery without having to close NWare then start it again?

I see this issue in NWare 1.2.6 when I open NWare then load a file then compile it and then try to deploy it. If all the Nions in the system were turned on and alive on the network when I started NWare then all will deploy OK, however if any of the nions were not present when NWare started, then turning them on/connecting them does not work, they will never appear in the deployment network view window at this stage.

To recover I have to first connect the Nions and make sure thay are on the network THEN close NWare (also making sure in the windows process list that pandad has exited, if not then close it manually), then restart NWare then reload my file then re compile and hopefully at deployment time I will see all the nions in the system, and can then deply successfully.

It would be very very useful to be able to force this discovery process from within Nware, as when working with lage systems/files much time is taken up by reloading files and re compiling just because one Nion was disconnected/powered off for some reason when NWare was started.



I frequently have problems where

2

Re: Nware NioNode device discovery

You should not be having this problem, Pandad is always discovering Nodes.

The discovery packet of "I'm a instance of Pandad, and this is what you need to talk to me" when a Node starts, or you first start NWare, goes out on the network.  This packet "bursts" on the network in the beginning, and is then quickly suppressed, going down to about 11 times a second (it is a tiny little packet, the whole process takes less than .9% of the 100mbps segment after startup).

It appears that you have read the Pandad Interactive Mode white paper, but for those that may want to read this, it is found here;

http://downloads.peavey.com/mm/Class_Ma … Primer.pdf

Tucan, what I would like for you to try the next time you have this problem, is to stop Pandad as a Process, and start Pandad Interactive Mode.  Then type "list" in the command line, and see if the missing Node is found here (look for the IP address).  If it is not found, make sure you can ping the Node.

If it is found here, but not in the NWare's Remote Log, then we have an undocumented bug that we need to figure out.

I would take a look at the network, and see if something is blocking the Pandad from the newly started Node from broadcasting it's "stuff".  I can't think what that could be, but perhaps other members of our forum can.

Let us know if this helps, or if there are more details that you can add.

Good luck.

Make it intuitive, never leave them guessing.

3

Re: Nware NioNode device discovery

Pandad is continuously trying to discover new NIONs on the network, however, it is possible to crash Pandad so it is still existing, but not working, as an "Image Name" on the "Processes" tab of Windows Task Manager.

Have a look at this document we recently put together on Pandad and how it works.

http://downloads.peavey.com/mm/Class_Ma … Primer.pdf

-Josh

Josh Millward
Burnt Orange Studios

4

Re: Nware NioNode device discovery

Josh, I am going to be away from real systems for a week, so will try some diagnosis that you suggest when I get connected to a real system again.

In the meantime however this is a problem that I have often found, so often in fact that whenever i get communications problems I expect pandad to be frozen (ie when i close Nware pandad remains in the process list, and if I open Nware again I get 2 instances of pandad), and therefore routinely check it in the process list expecting to have close it manually.

Having said that it now happens less in Nware 1.2.6 that I have been using heavily than previous versions especially 1.2.4 which I used for a long time.