BEFVP41 - ping and telnet problem

[spinge]

Here is my problem

I have a pair of BEFVP41's, and I can get them to connect fairly reliably (IPS problem - different story), but I am having a ping problem. The network is set up as follows:

Main Office: DSL (1.5/128)

IBM/AIX telnet server : 192.168.1.1
BEFVP41 : 192.168.1.6
PC's : 192.168.1.11 - 20
Linksys printserver : 192.168.1.100
HP printserver : 192.168.1.102

Remote Office: Cable (1.5ish/128)

BEFVP41 : 192.168.2.1
PC : 192.168.2.11
Linksys printserver : 192.168.2.100


There is currently only 1 PC at the remote office, but will be expanded once the system is up and running. When I sit at any PC in the main office, I can ping anything on the network (telnet server, PC's, BEFVP41, and both print servers), and can run a telnet session (inventory management system) from any of the PC's.

When I sit at the remote office, however, I can only ping 4 of the PC's, and everything else times out (even the BEFVP41!!)

I have netbios set up on both of the BEFVP41's (I have upgraded the firmware on both), I have set up RIP1 on both, and also defined a static route on each one.

I am able to set up shares from the remote office to the 4 PC's that respond to the ping, but they do not show up in the network neighborhood. That is not the big concern, though. The initial concern is to get the telnet up and running (the current connection is through a VERY old leased line running at 2400 bps, and into the serial port of the PC, so ANYTHING will be better). When testing, I disconect the serial connection, just so I know it is not part of my problem. I also have port 23 of the BEFVP41 in the main office forwarded to 192.168.1.1 (the server), otherwise the PC's are just on a peer to peer setup right now.

Any help will be greatly appreciated.

Rich

 

[markku]

Rip and static routing are not necessary, should work as gateway mode.

Is the gateway of the servers pointing to Linksys ( 192.168.1.6 ) in the main office ?

 

[spinge]

markku

Thanks for the tip on the RIP and static routing - I think I had read in one of the other BEFVP41 threads (the really long one) that someone had set up the routing and got their system to work - effectively that computers on network A did not know where to look for IP's specific to network B, and so it just went to the internet.

I think you are on to something with the PC's in the main office - the 4 PC's that respond to the ping are the PC's that have access to the internet, and their gateway address is set to 192.168.1.6 (the BEFVP41). (I should have caught that one - DUH)

As for the printservers (both the linksys and the HP) they are pretty much just "dumb" servers - "here is my IP, send me something to print". At least I think they are - I will need to check into them further to see if they can be enabled for that.

The IBM/AIX telnet server also does not have a gateway set up, it is also for the most part "dumb" in that it just responds back to the IP that requested telnet information.

That is where I was hoping the BEFVP41's would come in. The computer(s) on the remote end would make a request to the IP of the server, the BEFVP would recognize that it was on the main network, shoot it through the tunnel and then the server on the main end would grab it. When the server then responded to the request, it would see that the request came from the main office BEFVP41, send the information back to it, and back up the tunnel to the remote office and the requesting computer.

Is this the correct thinking? I am a bit of a newbie to the whole VPN thing. Or does the request from the remote network show up as the IP of the actual requesting computer (192.168.2.x) and therefore I would need to change the subnet mask on the IBM/AIX to accept requests from two networks? (something like 255.255.0.0)

Thanks again for the help

Rich

 

[markku]

Hi Rich,

VPN- tunnel is only routing the packets from network A to B, therefore the gateways should be defined, otherwise it will not work, do not ask how I know this.

The gateway tells the machine where to send the responses, otherwise the packets have no route back.

You can also define the gateway to the printer servers as well. Both networks should be defined correctly with local IP/netmask/gateway (=Linksys ) info.

 

[spinge]

Markku:

Thanks again for the help. I am not only the "Computer Guy" for our company, I also have other functions. I will get a chance to work on this a little over the weekend, and the beginning of the week, and I will let you know how I (We) did.

It sounds like you have run into this problem yourself, and learned the hard way, as I am.

Thanks

Rich

P.S. Is it true that the RIP protocol ends up sending alot of information across the network? I am thinking that removing it will help speed up the VPN. I understand that the VPN will be nowhere near the speed of the local network (I am limited to 128k upload on each end) but it is still a FAR cry from the current 2.4k serial connection. I simply want to keep the VPN as fast as possible. Thanks

 

[markku]

RIP is not necessary at all, just gateway mode. No other static routing, the boxes will do the routing automatically. Do not know about overhead of RIP.

Do not allow NETBIOS broadcasts, how attempting this may be, just connect \\remoteIP\sharedfolder -> less overhead.

All the best.