Recently our node optimization is mainly aimed at the general settings，which include the fixing the public net IP addresses, matching the public net IP addresses for the local area network as well as crossing the nets. The optimization of IPFS routing in this period concentrates on bootstrap list
For naming, we used the IPNS, which increased the Cache. But we didn’t find corresponding promotion of the quantity of node. Does it mean that the IPNS is just used to expose the content? Can the IPNS realize the optimized function in connection the net?
Besides, we still have the following work and puzzles needed to be solved.
Question1:how to understand conception of relay, and how to become the node of relay;
Question2, Random-Walk is a Discovery Protocol for DHTs, but Random-Walking is inefficient, we want to do something to speed up the process of discovery, for example, formulate a machine-processed by which can influence moving of routing, this machine-process may be the search of location. Would you like to share some suggestions?
Question3, since we have several nodes in different locations, we are planning to use parts of these nodes to build a private network including server nodes, home nodes as well as personal computers to test fluency of the private network of IPFS and the effect of absence of the http.
Question4: the gateway is not enough, if we want to build a gateway, Would you like to share some suggestions?
Question5, is it necessary to install the ngrok to linux when the intranet is complicated without public net IP?
Question 6,:do the randomwalk and mdns just affect the discovery and the swarming of the nodes? We expect to know whether they can be used in nodes transporting or whether they still need the help of route when transporting.
Question7:If we want to build a hot node through which more people can obtain data, Would you like to share some suggestions?