Two websites on different networks - two NICs

Hello,
I have got an xserve, running 10.4.3 with two NIcs on different networks (192.168.... and one on a 10.0.0...) , connected to two different ISPs. I would like to run two websites, one per ISP/network card.
I've read through Apple's doumentation - I think the problem is that OSX will only connect to the internet via one network card. Any I dea what I can try?
I have seen a post on the discussion board regarding the ipfw.conf and firewall
Thanks,
Nico

It has to do mainly with a routing table. In that routing table you tell the router if the destinition IP is X from Subnet Y send it through Gateway Z like this:
Destination | Subnet mask | Gateway
192.168.40.0 | 255.255.255.0 | local
10.65.1.0 | 255.255.255.0 | local
192.168.32.0 | 255.255.255.0 | 192.168.40.13
192.168.35.0 | 255.255.255.0 | 192.168.40.14
192.168.0.0 | 255.255.0.0 | 192.168.40.13
10.0.0.0 | 255.0.0.0 | 10.65.1.25
0.0.0.0 | 0.0.0.0 | 10.65.1.1 (Default)
Just set-up where things can come from and where they should go. Setup a good defauly and you should be set.
Hope that helps a little.
Sorry if the table is a little hard to read.

Similar Messages

Different afp shares for different network cards (NIC)

Hi,
I have a machine sharing via two connections on two separate physical connectors.
One is restricted to LAN, the other is allowed to use the internet connection.
How do I set up shares only available on the LAN connection?
Thanks

Sorry it took me so long to answer, I've been busy lately....I've found a solution to the problem I mentioned, and just in case anyone has similar problems in the future, I'll shortly describe what happened.
As Caseyw suggested, it was necessary to enable both network cards through the Measurement and Automation Explorer. The cause of the connections malfunction was actually that I wasn't using the "right" protocol for the Modbus communication, which ran on the secondary adapter. The solution was to use the URL protocol with the correct path on the field, addressing the right IP address. To avoid making this post a mixture of topics, I won't elaborate futher, but I got the gist of it, so if anyone is having similar problems whether it is working with several network cards or with Modbus communication protocols, feel free to contact me, I'll be glad to help.
Thanks

How to use two different network cards on RT?

Hi everyone... I'm currently working on a project where I have to deal with the issue mentioned in the thread's topic: I have a PC with RT LabVIEW that has to be able to establish network connections using two different network cards. One will be used to connect via TCP/IP with a host computer that will show the data transmitted through shared variables; the other will communicate with another PC through Modbus protocol. The key is that each communication is done through a separate network card.
So far I haven't been able to figure out how to configure both things to happen. Does anyone know how to do this? Any tips will help.
For the RT communication I'm using a standard RT project, with the RT PC being given one of the IPs, and for the Modbus part, I created an I/O server with a master and a slave. Separately everything works fine, but when I get them together it simply won't work.
Thanks for your help!
Solved!
Go to Solution.

Sorry it took me so long to answer, I've been busy lately....I've found a solution to the problem I mentioned, and just in case anyone has similar problems in the future, I'll shortly describe what happened.
As Caseyw suggested, it was necessary to enable both network cards through the Measurement and Automation Explorer. The cause of the connections malfunction was actually that I wasn't using the "right" protocol for the Modbus communication, which ran on the secondary adapter. The solution was to use the URL protocol with the correct path on the field, addressing the right IP address. To avoid making this post a mixture of topics, I won't elaborate futher, but I got the gist of it, so if anyone is having similar problems whether it is working with several network cards or with Modbus communication protocols, feel free to contact me, I'll be glad to help.
Thanks

My server has Two NICs ! I want to use one of the NIC for the internet and the other to a local Network .

In my case , there is a server ( windows 2008 enterprise) . Instead of sharing the internet using the server . i have done the following .
i have 24 port switch which is connected with the modem (ISP Gave it to me). My server then is connected with the switch . the client pcs are also connected with the switch(d-link switch).
my objective is to enable efficient way of file sharing from server for client pcs and also internet sharing at the same time for client pcs!
my server has two NICs ! how to make the local network and internet work with out a fight!
how could i achieve this ? help me guys
regards

It really depends on exactly what the equipment from the ISP is. If it is a simple modem there is no point in connecting it to a switch. It needs to connect to a router.
You can connect it to one NIC of the server and configure the server as a (NAT) router (with the other NIC connected to the switch).
If either of the switches is actually a router you can use it as the NAT router and run the server on the LAN using one NIC.
Bill

TS3367 FaceTime has a problem with calling (audio, video) between two devices when they are in the different networks.

Hi,
I have a problem with calling by FaceTime when me and my friend are in the different networks.
When i call to another device I see this call on that device and tap to the answer button after that i see a message "connecting" but the first device is trying to call in during about 1 minute and finish the call.
But if both of the devices have connected to the one network (wifi) it works correct.

Actually, it started working fine when I told it to capture
sound at 11kHz.. I should have tried that first, before making a
fuss here, sorry. Looks like some weird glitch, a driver problem
maybe, but anyway, switching to 11kHz seems to have solved
it.

IP Address with two NICs

LabVIEW 7.0
  I have Win2000, my client has XP Pro.
  My program, as a convenience, displays on the front panel the IP address of the machine it's running on. This is so he can enter this address in another machine and communicate with it.
  I'm using the STRING to IP function (input unwired), and the IP to STRING function to display the number.
  My client has TWO NICs installed - one for data acquisition (we are collecting data from Ethernet machines in a test cell), and another for general traffic (printing, inter-cell transfer, etc).
  We discovered recently that the program was displaying the local IP number (192.168.x.x), which is the DAQ net. This is not the one we want.
  So I found the MULTIPLE OUTPUT option on the STRING to IP function and turned it on. I used an INDEX ARRAY function to pick out address #1 (not #0) and display that. That works fine.
  My question is: how do I know that #1 is the right number to use? This app will be deployed on a bunch of machines, will the local be Net #0 on all machines? Is that a system config issue? Do I need to tell him to configure a control panel somewhere a certain way? Do I need to read both addresses and display the one that does NOT start with "192."? How do I make this bulletproof?
Steve Bird
Culverson Software - Elegant software that is a pleasure to use.
Culverson.com
Blog for (mostly LabVIEW) programmers: Tips And Tricks

Hi CMB,
There are several approaches to this. Some less favorable to others... such as at one extreme: to identify with a label which NIC goes to what network....
However, to keep automation... "automated", how about the following:
Since you can automatically detect each IP address, you can then implement a small test to "self-discover" which IP (NIC) talks tho which network. For instance, you may want to ping a known static address, such as a mail-server to discover that it is part of the WAN and not the test network. You may also prefer to ping by host name just in case the IP address of that machine changes.   In the event that you do not know any IP address or host names, you may try to ping something outside the LAN (over internet for instance), something that the test network could not reach (your own website for instance??).
Would you be autodetecting on an ongoing basis (ie. every day) or only during equipment setup and configuration?
Ray

WDS Transport on two NICs!

Hi! I'm configuring a server (win 2008) with WDS Standalone. The server does have two NICs, one is connected to a small home network (test environment) and the other is connected to a wide office network. Right now I do have DHCP set-up on the small home
network, and wds are configured to run over DHCP. The question I have is if it's possible to configure WDS to actually go through both NICs? I would like to have my test environment but still be able to PXE boot when at the users client too. The second network
does already have a DHCP server. Best Regards Kimzi

Hi Kimzi,
As far as I know the WDS can’t use the multiple NIC for the deployment at same time, if you want to deploy system to different subnet you can refer the following similar thread:
WDS PXE DHCP, Clients on different subnet
http://social.technet.microsoft.com/Forums/windowsserver/en-US/10998e36-c5cb-4548-9c68-0224217c118d/wds-pxe-dhcp-clients-on-different-subnet
I’m glad to be of help to you!
We
are trying to better understand customer views on social support experience, so your participation in this
interview project would be greatly appreciated if you have time.
Thanks for helping make community forums a great place.

RMI uses incorrect InetAddress on machine with two nics

Hello All,
RMI registry and server run on the same host. Host has two nics, one assigned for private network (192.168.x.x) and another for global.
For the sake of creating RMI registry I use the global one and specify it as four dot-delimited numbers instead of machine name:
Registry registry = LocateRegistry.getRegistry("xxx.xxx.xxx.xxx", rmiPort); // "xxx.xxx.xxx.xxx" is global IP addressInvoking a method with just a port as a single parameter causes RMI to bind to the private address, so I avoided that trouble by using numeric IP.
Remote client is able to locate the registry and extract the stub from it now, but ironically retrieved stub contains private address inside.
Debugger shows that the proxy stub contains TCPEndpoint which resolves to 192.168.xxx.xxx:yyyy
I am exporting the object on the server side by using UnicastRemoteObject.exportObject(myObj, 0) method.
Unfortunately, there is no method which would allow to specify the InetAddress to bind the object on.
I've seen the battle on page [Remote Method Invocation (RMI) - Why does RMI server sets its connection address?|http://forums.sun.com/thread.jspa?threadID=5366985].
Close questions are discussed, but I haven't found answer on my question.
Is there an ability to specify the address when exporting the object?
Or any workaround to let the client retrieve the proper stub which can be used to send requests to the server?
Thank,
Alex.

For the sake of creating RMI registry I use the global one and specify it as four dot-delimited numbers instead of machine name:
Registry registry = LocateRegistry.getRegistry("xxx.xxx.xxx.xxx", rmiPort); // "xxx.xxx.xxx.xxx" is global IP address
That doesn't create a registry. It just returns a stub to an existing registry.
Invoking a method with just a port as a single parameter causes RMI to bind to the private address.No it doesn't. It causes RMI to construct a stub using "localhost" as the hostname. If this maps to the private address via your /etc/hosts file or DNS, so be it.
Remote client is able to locate the registry and extract the stub from it now, but ironically retrieved stub contains private address inside.Nothing ironic about it. The stub doesn't know where you got it from. The address inside the stub is controlled by what "localhost" resolves to in the server JVM. If it resolves to something inconvenient, either change that or set the system property java.rmi.server.hostname to the desired IP address or hostname in the server JVM.
Unfortunately, there is no method which would allow to specify the InetAddress to bind the object on.You don't normally need one, see above, although in pathological cases you could use an RMIServerSocketFactory, which does give you that ability.
I haven't found answer on my question.See also the RMI FAQ, item A.1.
Is there an ability to specify the address when exporting the object? See above.

Mac Mini, two NICS and DNS

Hi everybody,
I'm not able to properly set up DNS role in Snow Leopard server with two nics. This is my situation:
NIC 01: WAN
Type: Builtin
IP: 192.168.1.252
SN: 255.255.255.0
GW: 192.168.1.254 (ISP Router)
DNS: 192.168.1.254
Domain: homenet.telecomitalia.it (ISP domain)
NIC 02: LAN
Type: USB
IP: 172.16.0.254
SN: 255.255.0.0
GW: empty
DNS: 172.16.0.254 or 127.0.0.1
Domain: mydomain.kzy
I'd like to provide DNS, DHCP, Directory service, print server with the LAN interface and use the WAN Interface only for internet access.Obviously all clients must have access to the web. In windows server there's a role called RRAS (routing and remote access) that allows sysadmins to do this quite easly, but in os x server I was not able to find nothing of similar. If i setup my server only with tha lan interface everything works great, but when I add the wan interface the dns stop working; seems that operating system confuse externale dns with internal dns. Obviously something is wrong in my configuration, please an anyone help me to solve this problem?
Thanks in advance
Alessandro

That's fairly typical of the problems encountered with OS X operating as an (expensive and slow and variously vulnerable) IP router. OS X Server can be configured as a gateway if you follow the set-up directions exactly.
It's local preference to configure either a commercial NAT-gateway-firewall-router box, or repurpose a spare two-NIC x86 box with M0n0wall, Smoothwall, pfSense or other open-source.
Preferably a box with a VPN server, but your 192.168.1.252/24 public-side IP address is going to cause issues with a VPN.
And in one reading of your configuration details, your ISP router is already a NAT-gateway-firewall-router box (based on that 192.168.1.252/24 private address you've shown), so you might want to investigate what its particular capabilities and features are. With that NAT assumption, you're headed toward double-NAT connection, and that'll play havoc with your networking. Single-NAT is problematic enough.
If that "ISP router" box is a NAT-gateway-firewall-router, then you'll be reconfiguring it for a subnet in your 172.16.0.0/12 or other private address space (which would make VPN pass-through more widely reliable), or moving your network into 192.168.1.252/24.
If you post the identity of the particular "ISP router", somebody might know its capabilities. Folks at your ISP almost certainly will, of course.

Two NICS presented to OS, possible to only present one ?

B200-M1 blades and Menlo CNA.
The SPs are using a dynamic vNIC templates with fabric A or B selected and F/O enabled.
What we're witnessing is the OS will always be presented with two NIC's. Depending on which Fabric A/B the SP is set to determins which NIC is active. eg : In Linux eth0 & eth1 are detected. If the SP points to A eth0 is active, and B then eth1 is active.
My question is : Using the Menlo card is it possible to present only one NIC to the OS and still retain f/o ?
We use a number of auto config tools and really want consistency on which NIC is active but also still want to use both A & B fabrics. Rather than assigning all SP' to A.
Paul

Paul
These are 2 static PCI devices (with device numbering) which the OS will show.
Default behavior with Menlo (linux terminology):
Lets say you don't create any vNICs in a SP and give it to a Menlo adapter based blade.
Your ifconfig should show eth0/eth1 while the link on both will be down (i.e unusable as the FI side config isn't done).
If you create 1 vNIC (side A), your eth0 will have link up with correct VLANs, eth1 down.
If you create 1 vNIC (side B), your eth1 will have link up with the correct VLAN, eth0 down.
It is for the way OS enumerates the devices. The first ethernet device is wired to side A.
You could potentially try changing the HWADDR/MACADDR in the /etc/sysconfig/network-scripts/ifcfg-eth[0-1] files to see if it works for you i.e assign an ip based on it.
With M81KR as I mentioned before you don't have this issue. You can call out the PCI order, number of devices and which vNIC goes to which side.
So 2 vNICs created in the M81KR could both be going to side A for example. This vNIC config will fail on a Menlo.
Is the use case that for a singly attached server, you are trying to do half the blades use Side A while the other Side B while maintaining uniformity in your imaging tools?
--Manish

Does a iMac mid-2014 have two NIC cards?

I was looking at my system info and noticed in Network > Ethernet one MAC address and then Network > Wifi > another MAC address.
I could only see one card Airport Extreme, I could not find info on this on apple.com specs or with a few lazy searches.
If I do have two NIC cards, How does that effect a loopback ping?

Do 'ifconfig' in a terminal window...loopback is on its own interface.

Lync edge and two NICs, DMZ and NAT

I am in the process of setting up my lync 2013 edge server and i seem to be stuck. We run a external firewall and an internal firewall and have a DMZ where all of my public facing servers sit that resides between these two I have things
like my webservers and edge transport for my exchange servers in here., I am in the process of creating my new edge pool topology and have a question. I am going to NAT my Edge servers from the external firewall so the public IP address is something
like 12.xx.xxx.xxx and then the DMZ subnet is 192.168.3.x. For my internal IP address I put the internal IP of my edge server which is 192.168.3.17 and for external setting i put the 12.xx.xxx.xxx IP address even though there is not a NIC in the edge
server that has that IP since it is NAT'd from my firewall. Will that work? The reason i ask is everyone seems to say that this edge server has to have two NICs that has one connected to my DMZ subnet and the other to my LAN subnet but doesn't
that mean you have a huge hole with a Windows server with one foot in the DMZ and the other in the LAN? Is htere a way to utilize only one NIC in an Edge server? I guess i am trying to see if it will work like the edge transport role in exchange.
thanks.

No, you shouldn't ever have an edge server with one foot in the DMZ and one foot in the LAN. That somewhat defeats the purpose of a DMZ since you've just created a path around your firewall. What you need is in effect two DMZs, one that communicates
only with the Internet, and one that communicates only with the internal network. This is where the two NICs come into play.
The external facing NIC would get private IPs typically (though they can be public) which are in turn NAT'd to public IPs. The topology builder knows about the private IPs, but for the A/V edge, there's also a section where you let it know that the
edge will be NAT'd and what the public IP for the A/V edge is.
You can get away with just one NIC, on just one subnet. People have had issues, and it's not supported, but I personally have gotten this to work without issue on several occasions.
So, if you simply can't have two DMZs, you could give the box one NIC and four IPs. 192.168.3.17, 18, 19, and 20. Assign 17 as the internal NIC, and 18, 19, and 20 as the access edge, web edge, and av edge. In the topology builder, specify
the 192.168.3.X addresses, but also put the 12.xx.xxx.xxx address in the public section. Put persistent routes on the box so that it knows to use the Internal firewall to get to internal addresses, and the default route should be the external firewall
(I suspect this is in place for other boxes in the DMZ unless they only talk to the Internet). Open your ports, add your DNS, install Lync and you should be good.
Please remember, if you see a post that helped you please click "Vote As Helpful" and if it answered your question please click "Mark As Answer".
SWC Unified Communications

Bonding two NICs

Hi,
I was wondering if it was possible to bond two NICs together in MacOS 10.4?
Thanks,
Andy

Server or Client?
In Server, for sure, but I've never looked to see if client supports it. Worth trying, I suppose:
System Preferences -> Network -> Show -> Network Port Configurations -> New
Here there's a list of interfaces you can add. One of them may be 'Link Aggregate' in which case, yes you can.
I take it you have a compatible switch to go with this?

Bridging two nics

Hello All !
So here is my problem:
Im running Dual PowerMac G4 as a server for small number of computers with some services and have two nics:
en0 with adress 192.168.0.1 (local network)
en1 with adress 192.168.1.2 (route to adsl modem with adress 192.168.1.1)
running NAT let people from local network surfing the internet and anything working fine.
but i can not bridging this two interfaces and create one with 192.168.0.1 adress for example. So how i can do it ?
Running OSX Server 10.4.9 btw
Thanks to all for help.
Anton.

Hmmmm...
Strange, but Thanks anyway
Anton

Is it possible to redirect to same port two NICs

Dear members,
I have to setup my Solaris10/Apache server with two NICs, then I want to redirect the 80 port to another port like 777 for both of NICs. I mean, all the NICs should listen the 777 port and redirect incoming packets to 80 port. I tried this, but I was never able to set it successfully for both NICs. But when I unplumb one of the NICs, it's ok. The other is redirecting. But I cant do this for both of them.
Any idea about this ? Please help.
Thank you.

user5401524 wrote:
I have additional findings.
After commenting /etc/defaultrouter and flushing the routing table, I'm able to reach the forwarded ports.
If I try from a server which is in the same subnet with this server I'm able to ping, telnet the port or reach via webbrowser. And the port forwarding is working too. But if a try from another subnet, it's not possible. It seems to me like a routing problem but I cant explain that yet.
Fyi.I don't know that the ipfilter syntax provided in the example is correct - I'd check with the ipfilter man page.
It looks like you are trying to redirect any outgoing address for port 7777 to 127.0.0.1 port 80, but when your packets
are looking up the routing table, and finding the default route first so that the outgoing packets would never
really have the 0.0.0.0 ip destination that your rule is set up for.
--Sowmini

Two websites on different networks - two NICs

Similar Messages

Maybe you are looking for