I have experienced a variation on this problem several times now..either an app that I have configured will not serve out on the machine that is serving it (though is it accessible from the outside world) or a site being hosted on a machine that is a part of our network (using a seperate IP, though still technically a part of our local network) will serve out to all outsiders, but not users on the network with the macihne.
In this particular case, I am running IIS 5 on a win2k machine that is not configured for DNS, DHCP, or Domain Control services...The server hosts a site that resolves outwardly using host headers (the dns record matches the host headers configured in IIS) and the site is capable of being viewed by 99% of the world. The issue is that all of the machines on our network (though on a differnt IP on that network) cannot access the site.
To be clear, the local server can browse the site, any outsider can browse the site, but all other users (on a segment of the network that uses a different IP from the web server, lIS is on .75 while networked users on .74 wan-side) are unable to view the site.
I have tried to ping and tracert the domain name on machines from the network that are unable to view the site, and all return the correct dns translation of the IP address as well as packets responding from the router that the server is connected to. The trace results in the same trace as that of the local web server (that can traverse the site content), but when one tries to browse the site, the request times out.
To date, my logic is as follows...firewalling being done by our local routers could be responsible, but this does not explain why outside traffic can get to the site. Also, the request to view the site is a vanilla http request packet being sent out using a host header value...so the path should be able to go out to the outside world and then back...but instead, it attempts to connect directly to the address of the web sever (I know this because tracert-ing the name results in one hop from both sources invloved - the local web server, that does serve the content and the local network machines that do not). So, at this point, I think that the packets are getting stuck at some point locally.
The topography involved is:
Router for Web Server & SQL Router for LAN Workgroup
| | | |
IIS Server SQL Server App Server / \
Workstations (10 or so)
The request from either the App Server or any of the Workstations results in a timeout, while the SQL & IIS machines both get to the site on the IIS machine.
If anyone has any clue as to why this could be happening, please help out with some suggestions...