Set proper messaging and expectations about MOC clusters

[tumido]

https://chat.google.com/room/AAAAAYpDdc8/_iy7t8NFV-I

[durandom]

This applies to the operate first environment in general.

test clusters

• can be destroyed by the requesting party at any time
• requesters can grant access to collaborators
• can be destroyed by operate first ops with X days prior notice

production clusters

• can go down anytime due to software failures
• no backup guaranty for data - but you can work with ops to implement a strategy for your workload
• scheduled outages are communicated with X days prior notice
• support is best-effort in a community fashion - no SLAs
• operations are practiced in a state of the art SRE fashion, aiming for maximum transparency

[tumido]

Transferred from moc-cnv-sandbox repo to ops-issues and added to the Operations dashboard.

We also need proper understanding (in written form, link-able on the website) which cluster serves what purpose and we should not expect users to deduce it from the cluster name or something.

[HumairAK]

@durandom -- given your definition, should moc-cnv-sandbox be considered under production or test? Given the discussions in the chat, it seems to me there's confusion in this particular area.

[tumido]

link of pull in the https://github.com/open-infrastructure-labs/operate-first-workloads into operate-first.cloud