I’m running js-ipfs in node.js, and I’ve noticed that my node stays connected to the Bootstrap nodes even when I repeatedly issue ipfs.swarm.disconnect() on them. Can anyone explain why that’s the case? Any suggestions on how to actually disconnect from a peer and stay disconnected?
Background
I’m still struggling with bandwidth issues. I’ve been able to isolate these bandwidth spikes to the Bootstrap nodes. In particular, these nodes:
- /dns4/node0.preload.ipfs.io/tcp/443/wss/p2p/QmZMxNdpMkewiVZLMRxaNxUeZpDUb34pWjZ1kZvsd16Zic (147.75.69.147)
- /dns4/node1.preload.ipfs.io/tcp/443/wss/p2p/Qmbut9Ywz9YEDrz8ySBSgWyJk41Uvm2QJPhwDJzJyGFsD6 (147.75.100.9)
- /dns4/node2.preload.ipfs.io/tcp/443/wss/p2p/QmV7gnbW5VTcJ3oyM2Xk1rdFBJ3kTkvxc87UFGsun29STS (145.40.65.83)
I’m running my own private network, complete with circuit relays and bootstrap nodes. I don’t need or want to use the public bootstrap nodes. I have the public bootstrap nodes in a blacklist. Every 30 seconds the node actively scans the ipfs.swarm.peers() endpoint. If the public bootstrap nodes are found in the results, ipfs.swarm.disconnect() is called on them.
However, using iptraf-ng (on Ubuntu), shows that my IPFS node is still sending massive amounts of data to the IP addresses associated with those public Bootstrap nodes.
I’m looking for a way to prevent the node from connecting to those nodes, as a possible solution to the bandwidth problem.
I’m using js-ipfs v0.55.4, as that’s the highest version compatible with OrbitDB, which I’m also using.