Hub Transport - adding second server for manual failover, in case of network outages

Hi,
You got to have a Hub Transport Server in every site where you have Mailbox Server role.
So if in every site you have one HUb role along with Mailbox role, and if that Hub goes down, then mail flow from that site would be affect no matter what. For redundacny you just add another hub role in each site and everything will be taken care of. You cannot ask Hub from other site to do mail flow for mailbox server in other site
Hope this helps
Thanks
Nitin Gupta

Hello, Actually this is only for site1 at this time. The VPN connection between Site1 and Site2 is on the same link used for internet by the cas/hub in site1, so the network outage took out both the vpn link between sites and the site1 cas/hub internet.

The other link I have is much slower than my primary link and I would like to make use of it only if the main link goes down again. I don't need automatic failover, so manually turning on a VM image is fine by me, I just don't know for sure if having the VM setup then turned off will cause problems and if everything will work fine with it off, then work fine if the main link goes down again and I power the VM on to basically point internet mailflow out the other link (as well as CAS) - it's not for load balancing or anything, just manual failover

Rather than creating VMs, you could:

assign a separate IP address to a "disabled" NIC
create the necessary send connector that use only that IP address (again, this should be in a "disabled" state)

When the main link goes down, enable the NIC, and enable the connectors.  I am assuming here that you can re-configure routing on your router/firewall as well.

Just talked to my boss - he wants to stick with VM - so it would basically be a backup HUB going out a different link. The server would be shutdown until needed - would this break Exchange?

I've not had any experience of creating an Exchange server and shutting it down until needed, so, sorry, I really don't know what would actually happen and how it would affect the Exchange organization.

My feeling is that it would not be a good thing, but that is unfortunately not a well-informed opinion.  It might very well not be problematic at all.  
 
The management tools will probably take a little longer to open as they search for the "missing" server.  And, you might find something about the missing server in the event logs. You will never again get a clean result when you run the Exchange BPA though  :-(

Hi,
2 HUb in same site and pulling down one, would not cause any issues in continuity of flow, but could cause issues if there are mails waiting to be relayed etc. Otherwise pretty much ok
Test it for yourself :-)
Thanks
Nitin Gupta (gupnit)

I believe I have convinced the execs that a 2nd disabled server in VM would not be the best way to go, as the server does have dual NIC. If I have the 2nd NIC disabled, as Kaffiend mentioned, is this correct?:

As mentioned earlier,  2nd NIC (disabled) has different IP and connected to VLAN that points out other Internet connection as well as send connector is created and disabled

Does this sound correct: if primary network goes down I would disable the primary NIC, enable the backup (do same with send connector), change internal DNS so servername points to new internal IP address and restart transport service?

thank you,
Kedar

Oh, and one more thing:

Might as well create a reverse DNS record for that other link now (have it in place before you need it)