1. create network with two organizations (each with one peer)
2. break network between them
3. wait "panic" because of
4. restart peer's containers
5. wait more then 1 hour (120 attempts to reconnect)
6. restore network between peers
FAB-15840 was fixed in v1.4.2, but we have the same problem on this version (please, see steps to reproduce)
I checked: https://github.com/hyperledger/fabric/blob/a282f7132565d509d3ba6d502cd9c07cc5e567f0/gossip/discovery/discovery_impl.go#L32
and found only 120 attempts to connect to anchor peers. And is that all? A peer gets isolated after that?
What should I do to fix this?