Bug #38425: Unable to run a rex job when a provisioned host is deleted and created again with same IP address - Foreman Remote Execution - Foreman

Actions

Copy link

Bug #38425

closed

Unable to run a rex job when a provisioned host is deleted and created again with same IP address

Added by Adam Lazik about 2 months ago. Updated 6 days ago.

Status:

Closed

Priority:

Normal

Assignee:

Adam Lazik

Category:

Target version:

Difficulty:

Triaged:

Yes

Bugzilla link:

Pull request:

https://github.com/theforeman/smart_proxy_remote_execution_ssh/pull/125

Fixed in Releases:

smart_proxy_remote_execution_ssh-0.11.5

Found in Releases:

Red Hat JIRA:

SAT-32007

Description

Description of problem:

When the connect_by_ip setting is set to yes, Foreman is unable to run a rex job when a provisioned host is deleted and created again with same IP address

How reproducible:

100%

Steps to Reproduce:

1. Provision a host
2. Run a rex job (set Connect by IP to yes)
3. Delete the host and create a new host(give same IP to the host as of the first provisioned host)
4. Try running a rex job again

Alternative steps to reproduce

1. Have the setting connect_by_ip set to yes
2. Register a host to Foreman
3. Run a REX job
4. Delete the host from Foreman
5. On the host run: `rm -f /etc/ssh/ssh_host_* && ssh-keygen -A && systemctl restart sshd && <registration command with force=true>`
6. Run the REX job again

Actual behavior:
The rex job fails on the host.
The job error is `Offending ED25519 key in /usr/share/foreman-proxy/.ssh/known_hosts:1`

Expected behavior:
The rex job should pass