Some more details, from
http://www.microsoft.com/technet/community/columns/cableguy/cg1105.mspx.
Improved Routing Path Detection and Recovery
To automatically attempt a new routing path for remote traffic when TCP
connections begin retransmitting segments, the TCP/IP stack in Windows
Server 2003 and Windows XP supports dead gateway detection. Dead gateway
detection is a fail-over mechanism that automatically changes the default
gateway of a computer to the next configured default gateway (when a
computer has been configured with multiple default gateways on an
interface). However, dead gateway detection in Windows Server 2003 and
Windows XP does not do the following:
* Distinguish whether the local default gateway has failed or a remote
gateway (router) has failed.
* Provide a fail-back method to change the default gateway back to the
primary default gateway when the primary routing path is restored.
The Next Generation TCP/IP stack addresses these issues through Neighbor
Unreachability Detection for IPv4 and fail-back support for an automatically
changed default gateway configuration.
Neighbor Unreachability Detection for IPv4
Neighbor Unreachability Detection is a feature of IPv6 that allows a node to
track whether a neighboring node is reachable. The neighboring node can be
either an on-link destination or a router. IPv6 tracks reachability through
an exchange of unicast Neighbor Solicitation and Neighbor Advertisement
messages or by relying on upper layer protocols such as TCP to indicate that
a connection is sending and receiving data. With Neighbor Unreachability
Detection, an IPv6 node can quickly determine when a neighboring node is no
longer reachable-due to node failure, link failure, or a node that changes
its media access control (MAC) address-and can take the appropriate steps
for recovery.
The Next Generation TCP/IP stack also supports Neighbor Unreachability
Detection for IPv4 traffic by tracking the reachable state of IPv4 neighbors
in the IPv4 route cache. IPv4 Neighbor Unreachability Detection determines
reachability through an exchange of unicast Address Resolution Protocol
(ARP) Request and ARP Reply messages or by relying on upper layer protocols
such as TCP. With IPv4 Neighbor Unreachability Detection, IPv4-based
communications benefit by determining when routers are no longer reachable.
For example, if TCP connections begin to retransmit segments, the Next
Generation TCP/IP stack tries to determine whether the current default
gateway is still reachable by sending a unicast ARP Request message. If the
default router does not reply, it is deemed unreachable and the Next
Generation TCP/IP stack changes the default gateway to the next one in the
list. If the default router replies but TCP continues to experience
retransmissions, dead gateway detection eventually changes the default
gateway to the next one in the list.
Fail-back Support for Default Gateway Changes
In TCP/IP for Windows Server 2003 and Windows XP, when dead gateway
detection changes the default gateway, the new default gateway remains the
primary gateway for default route traffic until dead gateway detection
switches to the next one in the list (cycling through the list of default
gateways) or until the computer is restarted. Therefore, dead gateway
detection in TCP/IP for Windows Server 2003 and Windows XP provides a
fail-over function, but not a fail-back function.
The lack of fail-back for default gateways can cause throughput problems on
a subnet containing two routers: a high-capacity primary router and a
lower-capacity backup router. The hosts on the subnet have the high-capacity
router as their first default gateway and the backup router as their second
default gateway. If the high-capacity router has a temporary failure, hosts
on the subnet will switch over to the backup router. When the high-capacity
router becomes available again, none of the hosts on the network use it
because they have switched to the backup router.
The Next Generation TCP/IP stack provides fail-back for default gateway
changes by periodically attempting to send TCP traffic through the previous
gateway. If the TCP traffic sent through the previous gateway is successful,
the Next Generation TCP/IP stack switches the default gateway to the
previous gateway.
In our example with the high-capacity router and backup router, if the
neighboring high-capacity router becomes unavailable, the hosts on the
subnet use Neighbor Unreachability Detection to switch their default
gateways to the backup router. The hosts then periodically attempt to send
TCP traffic through the high-capacity router. When the high-capacity router
becomes available and the hosts determine that TCP traffic sent through the
high-capacity router is successful, the hosts switch their default gateway
back to the high-capacity router.
Support for fail-back to primary default gateways can provide faster
throughput by sending traffic through the primary default gateway on the
subnet.
--
Steve Riley
(e-mail address removed)
http://blogs.technet.com/steriley
http://www.protectyourwindowsnetwork.com
Steven L Umbach said:
In IP properties advanced you can configure multiple gateways. HOWEVER
Windows can only use one default gateway at a time and the other is for
backup if one can not be detected.
You could however configure the routing table on a computer to send
specific IP/subnet traffic to a different gateway. You use the route print
command to see the routing table on a computer.
Steve
http://www.microsoft.com/resources/documentation/windows/xp/all/proddocs/en-us/route.mspx?mfr=true
--- using the route command
