Hi every one. 

Most of the Howtos on DMVPN configuration through OpenNHRP configure the ipsec.conf as follows:

spdadd gre -P out ipsec esp/transport//require;
spdadd gre -P in ipsec esp/transport//require;

My question is that if we are not specifying any specific ip address here for HUB <--> Spoke1 and similarly HUB <----> Spoke2, what about the dynamic ipsec tunnel that will be made between Spoke1 <---> Spoke2 on demand. I am uisng preshared secrets for tunnel formation. If Spoke1 is sharing the same secret with HUB, and similarly Spoke2 is also sharing the same secret, Spoke1 and Spoke2 can also form the tunnel without OpenNHRP or?. I am not understanding how phase 2 or 3 of the DMVPN is achieved with preshared secrets....

A spoke will specify the following configuration in its opennhrp.conf:

interface gre1 dynamic-map hub.example.com shortcut redirect non-caching

where hub.example.com will resolve to ip address of the hub.

What about the HUB opennhrp.conf? If no other HUB exist in the topology, whether the following file is ok?

interface gre1 shortcut redirect non-caching

Finally, the scripts given on the Alpine linux web site for OpenNHRP are only for alpine linux? , for example if we are testing in a lab environment with three Ubuntu VMs, do we need to run those scripts? because they also involve checking zebra running....