alias evmserverd as manageiq.service

[kbrock]

links:

• [ ] https://github.com/ManageIQ/manageiq-rpm_build/pull/408
• Alternative to #21396

All of our services go by manageiq-* and manageiq.slice. This allows us to start and stop services using the same name (i.e.: manageiq). It also allows us to use manageiq with journalctl.

I am suggesting introducing an alias instead of a rename to avoid breaking a single command: systemctl enable evmserverd. journalctl and the other systemctl commands will work the same in both cases.

Upgrading the rpm for an enabled service will not automatically create the alias. The linked rpm changes above adds the alias so the customer can use systemctl manageiq whether or not the service was enabled before the rpm upgrade.

Before

systemctl enable evmserverd
systemctl status evmserverd
systemctl disable evmserverd

After

systemctl enable evmserverd
systemctl status evmserverd
systemctl disable evmserverd

# added:
systemctl status manageiq
systemctl disable manageiq

Alt: Renaming the service

systemctl status evmserverd
systemctl disable evmserverd

# added:
systemctl enable manageiq
systemctl status manageiq
systemctl disable manageiq

# lost:
systemctl enable evmserverd

[kbrock]

@agrare I was looking around and not having much success. I do believe you are right that adding an alias to an already installed service will not add the alias. You'd have to disable / enable.

I do remember copying a file if a file already existed, but I did that a long while ago and can't seem to find the information

Some good information, but nothing I can use https://bugzilla.redhat.com/show_bug.cgi?id=1609065

Also wondering if we can merge before we solve the rpm issue.

[agrare]

Also wondering if we can merge before we solve the rpm issue.

As long as we have a plan for fixing the rpm issue and will be able to complete it in time. Otherwise there will be a difference in behavior for new vs upgrade. This isn't for petrosian so we have time but we need to do both.

[kbrock]

update:

• added alias fix / way to ensure the alias is setup in previous appliances.

[kbrock]

update:

• rebase

[miq-bot]

Checked commit https://github.com/kbrock/manageiq/commit/6b8da4e7cfd4bcb41caf1feb7cbbf4c5957b345d with ruby 2.6.10, rubocop 1.28.2, haml-lint 0.35.0, and yamllint 0 files checked, 0 offenses detected Everything looks fine. :+1:

[miq-bot]

This pull request has been automatically marked as stale because it has not been updated for at least 3 months.

If these changes are still valid, please remove the stale label, make any changes requested by reviewers (if any), and ensure that this issue is being looked at by the assigned/reviewer(s).

[kbrock]

notes:

We properly handle rpm upgrades for creating the links. I thought handling this was the reason for our blocker.

To be frank, even without this rpm change, the aliases seem like a good transition to using manageiq.service. Then we can update the documentation and move towards dropping evmserverd (long game not short game)