SpaceNetLab / StarryNet

A novel experimentation framework that enables researchers to conveniently build credible and flexible experimental network environments (ENE) mimicking satellite dynamics and network behaviors of large-scale ISTNs.
MIT License
64 stars 7 forks source link

Can StarryNet support routing researches based on reinforcement learning? #12

Open ZoeJiang2222 opened 2 weeks ago

ZoeJiang2222 commented 2 weeks ago

Thanks for all you have done! I wonder if there's any API provided for routing researches based on reinforcement learning? Or can you provide more related examples? Sincerely hope to get your help.

xiex386 commented 3 days ago

If you need APIs to modificate route tables, you can use Linux system APIs to apply static routes.

Iproute2 commands conveniently wrap these syscalls to interact with kernels. In containers, you can use the following command to operate route tables.

ip route <op> 10.0.4.0/24 dev B1-eth1 via 10.0.2.10, \<op> can be add/change/del.

With command to enter container, it will be:

docker exec ovs_container_1 ip route <op> 10.0.4.0/24 dev B1-eth1 via 10.0.1.10 in the main branch version or ip netns exec SH1O1S1 ip route <op> 10.0.4.0/24 dev B1-eth1 via 10.0.1.10 in the dev branch version.

ZoeJiang2222 commented 3 days ago

Thanks for your advice, I'll learn more about Linux kernels.