Simplify the alarm call to not re-query more them once.

This prevent launch activities for all objects in the list.

The amount of disable should be too small, so no need to worry so much here.

Archive Open Orders which contains no lines, there is no reason to keep a user active if he has no Hosting subscription

Don't set new stop dates smaller them the ones currently set. If the value is set for 3 months, we wait until that period come. This affects HS related to Subscription Requests, that required 3 months payment up front.

  If we do otherwise, and if the timestamp is propagated (send in another request), there is a chain of infinity
  recursive bangs

Label is hidden behind the panel.

Only consider slapos user who has at least one instance.

Now allocation goes into test.erp5.testSlapOSCloudAllocationAlarm

Upgrade Decision should be Cancelled if it is not applicable anymore.

This makes renaming easy before put the web page on a business template.

This is mostly to display Site Messages for various uses.

  Introduce initial logic for attention point

  For now we are using direct queries based on context, in future, we should
  replace it to also use Acknowledgements for a given context that are indirectly
  related. (Indirectly because Event is related to ticket, which is related to the Instance).

If computer all computers is down (red/no data), there is no meaning in calculate partition state, as all should be broken.

There are cases where you 'just' stop cron, so machines are acually reacheable, but I prefer consider it "broken" as the updates are not occuring on the machine.

slapos_cloud/SlapOSTestCaseDefaultScenarioMixin: Subscription allocation scope behaves like Public ones

This intend to reproduce the behaviour when monitor/upgrade based on allocation scope only.

Changes are related to the usage of monitor_scope definition.