When using policy routing, are the dynamic connected routes that get auto-generated by the system still applicable or do you have to manually specify the local interface values and gateways?
No answers on this? Is it because it’s a dumb question or just nobody knows for sure?
dynamic routes get added to the main routing table. if you do policy based routing you have usually additional tables ( with static routes ). with your rules/policies you select the other tables for specific traffic.
maybe you describe your setup a bit more in detail than it’s easier to give an helpful response 
Jaykay I think your answer is pointing me in the right direction. I just want to be sure I understand this correctly. If a packet gets marked and matches a rule and gets sent to a specific routing table, will the routing decision be made solely off of the static routes that you’ve created for that routing table? Since you said dynamic routes are only made to the main routing table, does that mean that once a rule hits a custom routing table that you have to manually recreate any entries to route traffic to specific interfaces and subnets that were previously dynamic routes? Yet another way of asking the question would be: In a custom routing table scenario for situations with dual WAN setups, if you want return traffic to be able to make it to the proper interface, do you have to manually recreate any routes that would normally be dynamic routes in the main routing table?
If you’re running dual WAN setup using policy routing you should not have any problem with the return traffic. So for example you have a policy to route voice traffic over WAN1. You phones sitting somewhere in the internal network on a LAN interface. The return traffic from the outside will be handled by the main table where you have all the routes back to you internal network. should not be a problem. Otherwise you need to tell us a bit about you setup.
My network setup is intense. If I go into major detail it will get messy in a heartbeat and I think over-complicate things. I have 6 sites each with dual WAN’s that I’m doing redundant IP/IP tunnels running over IPsec transport using policy routing for connection load balancing between each site/WAN link/VPN tunnel. The entire reason for my question is because I need to make sure that each connection traveling over each VPN tunnel returns via the same tunnel/internet connection with which it came from and this is where I ran into problems. I just need to fully understand how the routing tables work. After toying around with it quite a bit, I’m fairly confident the answer to all my questions earlier was yes but I guess I just wanted someone else’s re-assurance.