I have one LAN network with IP address 192.168.100.*** and other network with IP address 192.168.2.***. Every time I go to connect to each networked PLC I need to change IP address of my Laptop. Is there any way to set both IP address for network card in laptop. I know by using external or PCMK ethernet card I can connect to both network by changing to each network card. Is there any way to set both IP address in Laptop? And how two PLC with 192.168.100.1 and 192.168.2.1 can communicate over network.

Thanks and regards,

Shree

 

I have one LAN network with IP address 192.168.100.*** and other network with IP address 192.168.2.***. Every time I go to connect to each networked PLC I need to change IP address of my Laptop. Is there any way to set both IP address for network card in laptop. I know by using external or PCMK ethernet card I can connect to both network by changing to each network card. Is there any way to set both IP address in Laptop? And how two PLC with 192.168.100.1 and 192.168.2.1 can communicate over network.

Originally both networks 192.168.100.xxx and 192.168.2.xxx should be
separate. They just run over the same cable.

You could try to set the subnet masks to 255.255.0.0 on all PLCs and PCs.

Best regards
Friedrich Haase

 

Shree,

Dual ethernet cards are your only option to bridge two different subnets. You might try the utility
http://sourceforge.net/projects/changemyip/ I have found that it helps to quickly change your IP on a window$ machine.

Hope this helps.

Willey

 

Hi Shree,

It is very easy, I tell you the steps on Windows XP/200x.

1) Select your NIC from Network Connections panel ("Local Area Connection" by default).
2) Right Button click on your NIC and select "Properties"
3) Select "General" Tab and later, choose TCP/IP protocol and click on "Properties"
4) Choose "Use the following IP address" option and complete the info (IP and Subnet mask) to associate to your 1st network.
5) Select the "Advanced.." button later
6) In "IP Settings" tab you'll see the info about your 1st network which you just set. Now, click on "Add.." button and write the settings associated to your 2nd network and click on "Add" to save the info.
7) Later, click on "Ok", "Ok" and "Close" options and now you have access to 2 different networks no matter if they are class A, B or C.

Note: you can belong to N networks using this procedure.

Best regards.

Feliz noche!.

Gustavo A. Valero P.

 

Hello,

Is this technique of using two ethernet cards the way to deal with devices that are on two different subnets? I'm looking to move ethernet devices such as Cognex Cameras off of the company's main subnet and onto something dedicated to machine related things only. They are already on their own separate switches, but I feel that another level of isolation is necessary. We still want to use the computer that is on the main companies subnet to have access to the fileserver, etc, but still need to drill down to these other devices for programming on that same computer. Ideally I'd like to do this with one NIC.

Can a router be used for this if you don't want to have two separate NIC cards in your computer that programs your automation?

I know this is a basic question but any help would be useful. I've opened sockets and done lot sof programming over them to get the machine running, but it is still somewhat of a mystery how the information gets there.

Thanks,
KEJR

 

In reply to Ken Emmons Jr: I have used PCs with two Ethernet interfaces to isolate the machine I/O from the overall plant (office) network. I put the machine I/O on the built in interface, and connected to the plant network via a card I added for that purpose. Even the best cards are very cheap.

You just give each Ethernet adapter a separate IP address. This puts your I/O on a separate network from anything outside the machine. It also means that you can reuse the same IP addresses for I/O in each machine. This helps a lot if you have multiple identical machines, because every machine is configured the same. This means that if someone copies the configuration from one machine to another (which people will do), you won't have one machine talking to I/O that belongs to another.

If you just want to separate the plant from the office, then you can do this with router software or a proxy server. I can't recommend a specific solution for you, but have a look at Smoothwall, Clarkconnect, and IPCop. You probably only need a fraction of the features they offer, so the question will be which one is the easiest to set up and manage. They're all free, so it won't cost you anything to look at them. You can also do the same thing by working directly with IPTables, but learning to use that properly may take some time.

There are also dedicated router appliances (sometimes called "layer 3 switches"), but be prepared to pay through the nose if you want something flexible. If you decide to go this route, I would suggest talking to someone in your IT department for advice.

With a router, be prepared to spend a lot of time sorting out and labeling network cabling so that you can isolate the two networks correctly. It's one of these things that works well if it was planned that way from the beginning, but more difficulty than you can imagine if the network grew "organically", like in so many places.

 

Windows allows you to assign multiple IP addresses to a single NIC.

Its in one of the TCP/IP config tabs. It is only available if you have a
fixed IP address for your NIC.

 

In reply to M. Griffin:

What I'm trying to do is make sure that my industrial devices are guaranteed not to get hammered by network traffic. I have industrial Ethernet switches on the machine (physically mounted in the control cabinet), so traffic should only be going to the industrial devices that is meant for it. I was told that you want to put them on separate Ethernet segments by a former IT colleague, but was not explained why. I usually don't like that answer and want to know the reason, but was given the dreaded "Trust me" reply. The only reason I could see for going beyond a locally mounted switch is if there is some kind of broadcast error that floods every IP address on the network, like virus or other thing. Since our networking is not safety critical I would think this is OK in the strange event that it started happening and slowed the machine down (In this case I can always pull the plug and the machine will still run since the switch is local to the machine. We would just loose logging data temporarily until the network problem is fixed which is better than lost production...).

I will end up talking to IT regarding this issue, but I find it is preferable to know some approaches up front. A lot of the networking literature tends to be based on business related work and you have to be careful what assumptions you rely on. For instance, a 0.5 second latency might be [marginally] ok for a file save or complex database query, but would kill the machine cycle of a machine that runs faster than 1 second. I think in the long run that IT professionals will become trained more in the automation specifics, and vice versa.

KEJR

 

In reply to Ken Emmons Jr: There was a case a few years ago where the network for a nuclear reactor in the USA was knocked out by a virus (a "worm", technically). There was one computer on the network which used MS Windows. The computer was running SCADA software that was apparently logging safety related information (I think it was called the Safety Monitoring System).

The SQL Slammer worm got into it via a network connection that lead to the office, and from the office to their outsourced accounting firm (via an undocumented network connection).

SQL Slammer operated by taking over the MS SQL Server database program. It would then aggressively scan the network looking for more targets. The other (non MS Windows OS) nodes on the reactor network couldn't communicate because of network congestion. It was not immediately obvious to the plant operators what the problem was or how to fix it. It took hours of work to recover. Fortunately, the reactor was already shut down for unrelated reasons.

The point of that long story is that if you are worried about things like viruses (and worms), then you have to think about things like SCADA and HMI systems being the source of the problem, not just something that might be affected by it.

SQL Slammer would go out and actively search for anything attached to the network (this is what caused the congestion). A switch may or may not help you with that, depending on how smart the switch is and what is connected on the other side of it. Typical "industrial" switches aren't much different from consumer grade switches other than the packaging, mounting, and power supply options (they're just easier to mount in a panel).

You no doubt have heard that the "conficker" virus is expected to switch on new capabilities on the 1st of April. That will be an interesting one, because it has spread by things like USB keys (and taking advantage of an MS Windows feature that automatically loads it), not just network connections.

And yes, people in the automation field need to learn a lot more about Ethernet networking. It is really becoming part of the basic knowledge that will be required to operate effectively. I would also add things like http, html, and other web oriented things to that as those are infiltrating the field as well.

 

Excellent website for CISCO Routers including pictures of various router models all in one place.

http://routerpictures.blogspot.com

Regards.