Cluster Control SSH/Sudo Failure During PostgreSQL Cluster Deployment

Hello everyone,

I’m facing an issue while trying to deploy a PostgreSQL cluster using ClusterControl. The deployment fails with the following error messages:

[09:30:45]: SSH/sudo to host pg2 failed, job 44 can not continue.
[09:30:45]: Failed to access 1 host.
[09:30:45]: pg2: Can't connect (SSH): Socket error: Connection reset by peer.
[09:30:35]: pg2: Will retry in 10 secs (5/6).
[09:30:35]: pg2: Socket error: Connection reset by peer.
...
[09:29:54]: pg2: Checking ssh/sudo with credentials ssh_cred_job_6656.
[09:29:52]: Cluster will be created on 2 data node(s).
[09:29:52]: CMON version 2.3.3.14470.

The strange thing is:

I have already set up SSH key-based authentication.
From the ClusterControl server, I can log in to the database nodes without entering a password.
Despite this, when I start the cluster deployment, I still get the above SSH connection errors.

Has anyone faced this issue before, or can suggest what might be causing the “Connection reset by peer” problem during deployment?

Thanks in advance!

Permanently deleted user

September 02, 2025 09:37

Hello,

The issue here is that the connection is being forcibly closed by the remote host (pg2) during the initial handshake.

Can you please open a ticket with us? support.severalnines.com

Please make sure that you generate an error report by running this command on the clustercontrol node:

s9s_error_reporter -i 0

and attach the tarball generated in the ticket.

Regards,

Duncan

Dev Ops

September 02, 2025 18:17

Thanks a lot, Duncan! I’ll open a support ticket and attach the error report as you suggested 🙏

Cluster Control SSH/Sudo Failure During PostgreSQL Cluster Deployment

Comments

Didn't find what you were looking for?