enabling ceph image replication: how to set up host addresses?

lifeboy

Renowned Member
I'm attempting to do a test to replicate a ceph image to a remote cluster by following this HOWTO. However, what I'm missing is the detail of how or where to specify where "site-a" is in the examples given in terms of ip address.

When I follow the instructions, I see this in the status logs:

Code:
root@pmx2:/etc/pve# systemctl status ceph-rbd-mirror@rbd-mirror.pmx2.service
● ceph-rbd-mirror@rbd-mirror.pmx2.service - Ceph rbd mirror daemon
     Loaded: loaded (/etc/systemd/system/ceph-rbd-mirror@.service; enabled; preset: enabled)
     Active: active (running) since Wed 2024-06-19 11:11:53 UTC; 2min 3s ago
   Main PID: 3376697 (rbd-mirror)
      Tasks: 30
     Memory: 16.3M
        CPU: 374ms
     CGroup: /system.slice/system-ceph\x2drbd\x2dmirror.slice/ceph-rbd-mirror@rbd-mirror.pmx2.service
             └─3376697 /usr/bin/rbd-mirror -f --cluster ceph --id rbd-mirror.pmx2 --setuser root --setgroup root


Jun 19 11:12:23 pmx2 rbd-mirror[3376697]: 2024-06-19T11:12:23.381+0000 76efa2a75a80 -1 rbd::mirror::PoolReplayer: 0x64c3dc607800 init_rados: error connecting to remote peer uuid: d268da59-13d3-4f49-94d3-8b046b6571df cluster: capetown cli>
Jun 19 11:12:53 pmx2 rbd-mirror[3376697]: 2024-06-19T11:12:53.388+0000 76efa2a75a80 -1 rbd::mirror::Mirror: 0x64c3dc8afc20 update_pool_replayers: restarting failed pool replayer for uuid: d268da59-13d3-4f49-94d3-8b046b6571df cluster: cap>
Jun 19 11:12:53 pmx2 rbd-mirror[3376697]: unable to get monitor info from DNS SRV with service name: ceph-mon
Jun 19 11:12:53 pmx2 rbd-mirror[3376697]: 2024-06-19T11:12:53.431+0000 76efa2a75a80 -1 rbd::mirror::PoolReplayer: 0x64c3dc607800 init_rados: error connecting to remote peer uuid: d268da59-13d3-4f49-94d3-8b046b6571df cluster: capetown cli>
Jun 19 11:13:23 pmx2 rbd-mirror[3376697]: 2024-06-19T11:13:23.439+0000 76efa2a75a80 -1 rbd::mirror::Mirror: 0x64c3dc8afc20 update_pool_replayers: restarting failed pool replayer for uuid: d268da59-13d3-4f49-94d3-8b046b6571df cluster: cap>
Jun 19 11:13:23 pmx2 rbd-mirror[3376697]: unable to get monitor info from DNS SRV with service name: ceph-mon
Jun 19 11:13:23 pmx2 rbd-mirror[3376697]: 2024-06-19T11:13:23.488+0000 76efa2a75a80 -1 rbd::mirror::PoolReplayer: 0x64c3dc607800 init_rados: error connecting to remote peer uuid: d268da59-13d3-4f49-94d3-8b046b6571df cluster: capetown cli>
Jun 19 11:13:53 pmx2 rbd-mirror[3376697]: 2024-06-19T11:13:53.496+0000 76efa2a75a80 -1 rbd::mirror::Mirror: 0x64c3dc8afc20 update_pool_replayers: restarting failed pool replayer for uuid: d268da59-13d3-4f49-94d3-8b046b6571df cluster: cap>
Jun 19 11:13:53 pmx2 rbd-mirror[3376697]: unable to get monitor info from DNS SRV with service name: ceph-mon
Jun 19 11:13:53 pmx2 rbd-mirror[3376697]: 2024-06-19T11:13:53.535+0000 76efa2a75a80 -1 rbd::mirror::PoolReplayer: 0x64c3dc607800 init_rados: error connecting to remote peer uuid: d268da59-13d3-4f49-94d3-8b046b6571df cluster: capetown cli>
lines 1-20/20 (END)

If course it's clear that pmx2 doesn't know where "capetown" is, so how do I tell it?
 
The file rbd-mirror-capetown.conf contains the config of the capetown cluster on the remote cluster, so from that I assume that I have to create a VPN link between the two sites so that the replication service on the LAN at the remote site is able to get to the local / LAN address given in that aforementioned conf file. Is that correct?