Upgrade without reboot - how to know when to reboot after?

I’m trying to write an Ansible script to update my Github Enterprise nodes, and I do this 

# This command will reboot the node
- name: "Upgrade {{ github_host }} with version {{ github_version }}"
  command: "ghe-upgrade {{ pkg_dir }}/{{ pkg_name }} --yes"

You don’t have to understand Ansible to figure out what the above code does, which simply runs the ghe-upgrade command on the node. The problem is, the above command causes a reboot, and hence, I lose my ssh connection to the node (Ansible uses ssh).

I see that the ghe-upgrade command has a “–skip-reboot” option. So I issued the command (manually for now) with the option, and rebooted the system. However, when the system recovered, Github was not running at all.

What else do I need to do if I use the “–skip-reboot” option?

Hi @zillag,

We generally do not recommend automating GitHub Enterprise upgrades as these are intensive procedures and may run into problems.

However, doing an upgrade with --skip-reboot and then rebooting the appliance should work. Since your GitHub Enterprise appliance did not come back as expected, I would recommend contacting GitHub Enterprise Support and sending them a support bundle.

Thanks!

Thanks!

The “–skip-reboot” option worked. I just had to reboot twice, once for Github to finish its configuration, and the other for the configuration to take effect.

I’ll reserve my comments on a company that discourages automation for their product.

Hi @zillag,

This is one of the few areas where we discourage automation because of how it can impact users in a dramatic fashion if things go wrong. We have been looking to continue improving the upgrade process, so I’m happy to pass on any feedback that you have to the product team.

Thanks!