Fabric Version - 2.4.
TLS is enabled both for ORDERER_GENERAL and ORDERER_ADMIN.
3 EtcdRaft orderers configured, each with it's own admin.
Using fabric-tools, I can ping the administrator's host normally but when I'm try to run osnChannelJoin with ORDERER_ADMIN TLS cert and key - i get:
Error: Post "https://orderer1-admin:10443/participation/v1/channels": dial tcp 10.xxx.xx.xxx:10443: connect: connection refused
Error: Post "https://orderer2-admin:11443/participation/v1/channels": dial tcp 10.xxx.xx.xxx:11443: connect: connection refused
Error: Post "https://orderer3-admin:12443/participation/v1/channels": dial tcp 10.xxx.xx.xxx:12443: connect: connection refused
Maybe someone is confused with this behavior and can tell me what to look for? I will be very grateful. Thanks!
I tried to install the same host for the admin and orderer and refer not to the admin's address, but to the address of the orderer itself only with a different port.
Also, for the registering TLS admin certificate, I added the orderer host to the CSR hosts.
I also play with ORDERER_ADMIN_LISTENADDRESS - trying to set 0.0.0.0:10443, 127.0.0.1:10443. Nothnig helps.
It seems like all you need, for this kind of issue, is to use:
ORDERER_ADMIN_LISTENADDRESS=0.0.0.0:$ORDERER_ADMIN_PORT
And don't forget to expose $ORDERER_ADMIN_PORT in orderer container.