(see the attached network diagram)
You're network is connected to a vendor's network. You have a primary and secondary link to the vendor network. The secondary link is connected to the vendor network through your secondary site. You are running EIGRP for your main network (primary and secondary locations). You would like to inject static routes to the networks you need to access in the vendor's network. You are thinking about also using a floating static on your primary and secondary routers connected to the vendor.
Any suggestions for the best approach?
I think we still may be talking at cross purposes.
If you use floating statics on the secondary router the routes will NEVER be used unless the primary fails. An example -
ip route 220.127.116.11 255.255.255.0 10.99.0.68
ip route 18.104.22.168 255.255.255.0 10.99.0.68 250 <--- Notice the 250 here - this is the AD (Adminstrative Distance)
On both routers you add this under your eigrp config
router eigrp 1
So why don't you need metric's ?
The primary router redistributes the above static into EIGRP. When a route is redistributed into EIGRP it gets an AD of 170.
So the secondary router receives a route from the primary for 22.214.171.124/24 with an AD of 170. It also has a locally configured static route with an AD of 250. So it uses the route received from the primary router. It's own locally configured static is not used, not installed in the routing table and therefore never redistributed.
So the secondary router only has one route to 126.96.36.199/24 and that is via the primary router.
So this is why you don't need metrics because the only active route for 188.8.131.52/24 is via the primary router.
If the primary router fails then it stops advertising the 184.108.40.206/24 network. Because the secondary router no longer receives the route it now uses it's locally configured static to 220.127.116.11/24 which has an AD of 250. It's installs this in it's routing table and then redistributes it into EIGRP. So now all devices in your LAN see the path to that network via your secondary router.
If the primary router comes back up it's static route will be used again because it has a better AD.
Does this make sense ? Is this not what you tested in your lab ?