Alternate IP Address causing Domain connections to drop

This problem is a little tough to explain, but I'll try. I'd like to find if there is a resolution for it.

The way our company network is set up, we log into a domain. This gives me access to a number of network resources.

As a programmer of embedded equipment, one of the devices I've created can communicate via Ethernet to host systems.

Our network was originally laid out as a 253 system segment at 192.168.10.x (mask of 255.255.255.0) and this network segment has no free space for me to add a number of these controllers onto it since it's filled up almost to it's entirety with systems and servers.

So, I've set these units as addresses in 192.168.0.x, and I've created an alternate IP address on my system as 192.168.0.250, with a mask of 255.255.255.0.

This works, as the IP stack knows which address to send requests out to. If I address something in the 0.x range, the system originates at 0.250, anything else goes out as 10.x. The problem is, my network resources on the domain disappear occasionally, and when I delete the alternate IP address, they come back. I've been using this scheme for quite some time on Windows 2000, and it was never a problem, so something has changed over the years. It seems to work for a while, and then randomly the resources drop out.

Does anyone know why doing this would cause the connection to network resources to just vanish? It would seem that what m doing should work, since it does work. I just need to find the "irritant" that causes the connections to be lost.

Before you suggest we expand our network to support more hosts on the subnet, our IT department has rejected this. They don't want to have to go to every machine in the network (multiple buildings), reset subnet masks on clients and servers, and hope "it all works" after that.

Sarah,

The server(s) on the network are all 2003. The AV is McAfee VirusScan Enterprise 8.7. Other devices on the subnet the domain is on are the servers, other PCs, print servers. If I connect one of my embedded devices on that subdomain, no problem. It's only when I add an alternate IP to my PC that I loose the network resources.

There are no routers involved in this connectivity scenario. We have only one router leading to the internet, and these connections would not go through that. In fact, when I loose the network resources when the alternate IP is in place, I still have internet connectivity, so I don't loose connectivity to the router, only to the devices on the domain.

Hope this helps. Thanks.

Jeffs said:

This problem is a little tough to explain, but I'll try. I'd like to find if there is a resolution for it.

The way our company network is set up, we log into a domain. This gives me access to a number of network resources.

As a programmer of embedded equipment, one of the devices I've created can communicate via Ethernet to host systems.

Our network was originally laid out as a 253 system segment at 192.168.10.x (mask of 255.255.255.0) and this network segment has no free space for me to add a number of these controllers onto it since it's filled up almost to it's entirety with systems and servers.

So, I've set these units as addresses in 192.168.0.x, and I've created an alternate IP address on my system as 192.168.0.250, with a mask of 255.255.255.0.

This works, as the IP stack knows which address to send requests out to. If I address something in the 0.x range, the system originates at 0.250, anything else goes out as 10.x. The problem is, my network resources on the domain disappear occasionally, and when I delete the alternate IP address, they come back. I've been using this scheme for quite some time on Windows 2000, and it was never a problem, so something has changed over the years. It seems to work for a while, and then randomly the resources drop out.

Does anyone know why doing this would cause the connection to network resources to just vanish? It would seem that what m doing should work, since it does work. I just need to find the "irritant" that causes the connections to be lost.

Before you suggest we expand our network to support more hosts on the subnet, our IT department has rejected this. They don't want to have to go to every machine in the network (multiple buildings), reset subnet masks on clients and servers, and hope "it all works" after that.

Windows have a service that will broadcast onto the network to find resources (I forgot the name of the service), that service is updated in Vista and 7, when back in 2k and XP days, that service (I'll call it the "Browser" service from now on) sometimes can find resources, sometimes can't on a LAN, all Windows client. But when you add a Windows Server to the mix, somehow the "Browser" service runs perfectly fine.

Now, on the Vista and 7 generation, that service got updated big time. Now it can find ANYTHING with SMB protocol on the network, but for some reason it's "flaky", sometimes it might loose some resources and to list it few minutes later. I have my certification in Win2k+XP client and 2k/2k3 Server environment, I had my share of nightmares. All you can do now is just not rely on the "Browser", ie. don't trust what you don't see on the network browser, since the info in there is supplied by the weird service I mentioned earlier.

Either remove the second IP address, OR you can just ignore the missing network resource and access the host serving the resource by FQDN or direct IP address when it gone missing.

zzz2496

OK. A lot of replies to sift through.

1) It appears what I am doing is creating two Class C networks from my PC. One on 192.168.10.x, the other on 192.168.0.x.

2) The login to the domain is on the 10.x subnet.

3) "other" foreign network "Stuff" is on the 0.x subnet.

4) Only my PC needs to access the items on the 0.x subnet, but my PC also needs to stay connected to the 10.x subnet for internet access, Outlook access, and other network resources (storage, etc)

5) When this situation occurs, Internet access continues to function. Only the network resources that I need to log in to access disappear.

Either remove the second IP address, OR you can just ignore the missing network resource and access the host serving the resource by FQDN or direct IP address when it gone missing.

It's hard to ignore the missing resources when any program wanting them says the network is disconnected. As for the other side of your "or" I don't know what FQDN is or how to access those resources by direct IP (plus the resources are spread over multiple server IPs).

Removing the second address is what I want to avoid as I need to be able to access the control machinery on the network, but once again I can not put it on the 10.x subnet due to lack of available IP addresses in the 10.x subnet.