Hi,
I am trying to set a replication task between my local and remote machine but am getting this error when I select the destination path.
I have successfully set the SSH Key and connection via GUID but still getting this error.
What could be the reason for this? How to troubleshoot it?
Note: I have a firewall on my VPN and I have SSH and GUID ports open, it’s necessary to open an additional one?
Just to update that I forced the replication task creation and got this error:
Replication “concorde/docker,…,merlin/rpi_backup - galaxy/belgrado-backup” failed: Host key for server ‘100.117.219.87’ does not got ‘XXXXX’, expected ‘YYYYYY’…
I manually set the SSH Keys between both machines, added the authorised keys to the admin user on the remote and allowed to use sudo without password.
Still getting this error and does not know how to solve it.
Anyone could help me how to troubleshoot this?
Can you screenshot the whole replication window?
Have you selected your SSH Connection (top right) in that window?
Have you ticked both boxes for the admin user to:
Allow all sudo commands
Allow all sudo commands with no password
Are you doing PUSH or PULL replication? This depends on which admin user requires the above.
Here is the screenshot.
Yes, I have tried to set it manually and semi-automatically. Still the same problem.
Yes, I checked everything in the remote machine admin user. I believe that this was set by the semi-automatic SSH remote configuration.
I’m PUSHING from my local server to my remote server.
Ok so here is what I do.
This works for me.
Followed everything well less point 4.
Looks like the connection is being made but the replication is not.
If I set the destination path manually and continue then it gives this error [EFAULT] Authentication failed. which is not good. When I try the semi-automatic authentication I get the error Host key for server ‘100.117.219.87’ does not got ‘XXXXX’, expected ‘YYYYYY’…
Ok why not try with the root user or create another user. Something clearly upset somewhere.
Yeah, something very strange. Changed to root user and the problem persists.
Are you on the latest version of SCALE?
Yes, version Dragonfish-24.04.2 is on both machines.
Same problem here. Hope someone has found a solution but so far google is useless
Hi,
Yeah, I solved it from my side. It was my fault, I messed up with IPTABLES on my Tailscale Jail.
The problem was that my Tailscale Jail was forwarding everything from port 22 to the host server (I did that to access my Truenas via the Tailscale network), making the destination error Public Key because was pointing to my local instead of the remote one.
Not sure if this is your problem, but confirm that you are pointing to the right machine. I started to look at every Public Key on all my machines and found that was my Jail pointing to the wrong server.
I had the same issue with setting up a PUSH with a non-root user, although I was able to get it to work with the root user.
The source machine in this case is a relatively new installation, first installed with 24.10.1 and upgraded to 24.10.2. The destination is older, installed with 22.12.0 and steadily upgraded to 24.10.2.
I saw this post that mentions there were some replication issues with non-root users prior to 22.12.1. I wonder if this could somehow be related to issues originating from earlier versions and unresolved by upgrading.
Interestingly, I am able to set up a PUSH from the old machine to the new machine with a non-root user on the new machine; maybe a clean installation is the key here… just a guess.
I encounter similar problem, and for me was caused by a bad ACL on the home folder of the non-root user i was using for the replication.
But if i remember well i was doing the opposite > PULL from target and not PUSH from source
