hardice501
commented
1 year ago update issue in my cases, add option -p 30303:30303/udp is working at linux.
but in mac(m1), must need to build a virtual environment to run the docker cli.
- paid license(docker desktop, OrbStack) is working at M1
// the other are free license
- using program based on lima (lima, colima, rancher desktop): udp port is not enable
- using podman: p2p connection is possible only with boot_node. (If not specify MAC address, podman does not send the host's ip, but the url for the virtualized ip.)
in boot_node
> admin.peers
{
firstEnode: {
localhost: [virtual ip]:30303,
remotehost: [virtual ip]:{random port},
},
secondEnode:{
localhost: [virtual ip]:30303,
remotehost: [virtual ip]:{random port},
}
}so each peers(ex: first and second p2p connection) cannot connect.
- using multipass + docker-cli: can find peer(but must be assigned two physical IPs.)
non-fungible-nelson
fab-10
Description
We are Test Besu 21.10.9 Docker, IBFT2 consensus protocol, on a private network, gas price free with 4 validators.
4 Validators on 1 machine is no problem.
But build docker on each of 4 computers with different ip, bootnode cannot found each peers.
Acceptance Criteria
Steps to Reproduce (Bug)
Expected behavior: [What you expect to happen] Successfully connected peers Successfully Ping test Actual behavior: [What actually happens] No connection between peers. Successfully Ping test Frequency: [What percentage of the time does it occur?] Always except when adding the --net host option Ping test always succeeds even without --net host option
Logs (if a bug)
Please post relevant logs from Besu (and the consensus client, if running proof of stake) from before and after the issue. FullSyncTargetManager | No sync target, waiting for peers: 0
Versions (Add all that apply)
vmware -v]Additional Information (Add any of the following or anything else that may be relevant)
config option (if not boot_node, include option --bootnodes=${BOOT_NODE_ENODE}):