Can run FP using the second NIC card?

[Ben]

Hi Aaron et al,

I found myself in one of those rare situations where the customer actually asked me what type of hardware he should buy BEFORE he bought it.

Along with a list of memory CPU etc, I told him we should buy two NIC's in each PC.

What I wanted to do was ensure that the ethernet that will be used to connect my FP-1600's to my PC, was not sharing bandwidth with all of the other goofy things that could be going on at this facility.

So now the question:
I am using the latest and greatest FP VI to open, read (repeatedly in a loop) and then close when I am done.

When all of this hardware comes in, am I going to have a hard time getting the FP explorer to find the FP-1600's on the second NIC?

Have I un-nesse
sarily complicated my life?

Thanks to all of you for all of the entertaining and enlightening reading you have provided in the past.

Ben

[Aaron_G]

Ben,

Yes, you have complicated your life. The FP-1600 will work on computers that have properly configured dual ethernet ports (must be configured for mutually exclusive subnets). However, due to the combination of a bug in Windows and a bug in the OS that is in the FP-1600, the initial configuration of IP Address/MAC Address/Time Server is more complicated. There are several KnowledgeBase articles on this topic for your further and in-depth reading enjoyment, but it basically boils down to the fact that I would recommend you do the IP Address assignment from a computer that has a single ethernet address and then move the FP-1600 to the private subnet. You will have no difficulty talking to a configured FP-1600(reading values/writing values/configuring IO
moduless) from a Dual NIC machine, but the configuration of Ethernet settings may be problematic.

Regards,
Aaron

[Ben]

Cool!

So once I am passed the config of the ethernet I am good to go.

In the future, would you suggest I take the same approach or should I use a switch or router to issolate the FP traffic from the rest of the world?

Thanks again,

Ben

[Aaron_G]

Ben,

In the future you can take the same approach or use switches, but I honestly have not seen many cases where it makes all that much difference. I know of some people that have used routers successfully, but also a number who have not been able to get their routers to properly work with FieldPoint Ethernet modules.

In some testing that we did a few years ago, (I think it was back with firmware 0200 for the FP-1600 modules), we determined that a FP-1600 on a 10 baseT link to a computer (crossover cable) would generally use up to 7-8 % of the total bandwidth when it was streaming data as fast as it could (the bank had a large number of counter and FP-AI-100 modules). Adding a second bank increased the used bandwidth to about 14-15% if I re
call correctly. Remember, the FieldPoint Ethernet modules use a custom ethernet protocol designed by National Instruments. The protocol is designed for report by exception, which minimizes network traffic. What this means is that signals are transmitted only when they have changed outside their deadband range, unchanging signals do not result in any substantial network activity. Modules that sample relatively slowly (E.G. FP-TC-120 at 1.13 seconds) do not generate a substantial load on the network.

Regards,
Aaron

[Ben]

Ok.

My main concern is the other way around.

I did not want the normal network traffic to get in the way of the FP traffic.

The last thing I needed was to have some guy on third shift playing interactive video games to get in the way of my measurements.

Take care,

Ben

[pittpanther]

I have gone the same route as Ben (dual NICs to segregate the corporate LAN from the DAQ LAN). I was wondering if there are any special considerations when choosing the IP addresses for the FP-2000s?

Thanks!

[Aaron_G]

When configuring a second ethernet port, the subnet's must be mutually exclusive, there can be no overlap in IP addresses within the subnet. Remember that the subnet is determined by logically AND'ing the subnet mask with the IP Address. I have seen several cases where people used a 255.255.0.0 subnet mask and a 255.255.255.0 subnet mask and thought that it was exclusive...

If you are creating a private network on the second port that will not be connected to the internet, there are several recommended IP Address ranges. These are listed in the Appendix of the FP-20XX Series User Manual.

Regards,
Aaron

[pittpanther]

Thanks Aaron. I understand the notion of mutual exclusivity (in general!), but am having trouble applying it to this case.

The corporate network operates with a DHCP Server. Currently, the box is setup with a subnet mask of 255.255.252.0 and an IP address of 128.244.194.240. When I AND these together I get 128.244.192.0 as the subnet. Where does this leave me? Sorry for the questions!

[Aaron_G]

Okay, here's an example of mutually exclusive subnets. Your primary subnet (using the above numbers) is 128.244.192.X where X is any number between 0 and 255. All computers that have subnets of 128.244.192.X are considered part of the same subnet. Assuming that the second port network is not going to connect to the internet (e.g. private network), we will use an IP address of 10.0.0.1 and a subnet mask of 255.255.255.0. In this case the second subnet is 10.0.0.X and there is no overlap in IP Addresses between the two subnets. However, if the subnet mask for the first port was 255.255.0.0, then the subnet would be 128.244.X.X. In this case, we can not choose an IP Address of 128.244.193.2 since there wo
uld be overlap of the addresses.

Hope this helps clear it up,
Regards,
Aaron

[pittpanther]

Thanks Aaron - that definitely clarifies some questions I had. In this reference, when looking at the two subnets in binary - does mutual exclusivity imply that neither subnet shares a '1' in the same bit position?

BTW, I did define the FP-2000's IP as 10.0.0.2 with a subnet of 255.255.0.0, and was able to fully configure it from the dual NIC machine. This is my first exposure to FieldPoint and am so far very impressed. I am using the OPC Server to serve data up to Intellution's iFIX OPC Client. Within a couple of hours, I had everything setup and working! Thanks again for your help.