Description of problem:
In order to provide partners and clients with the tooling and mechanisms to successfully run the update/upgrades workflows there is the need to track the progress on the following sub RFEs:
Allow operators to run a minor update in OSP14.
* [RFE] Document the processes and workflows for the operator to follow when running a minor update in OSP14.
Allow operators to run a major upgrade from OSP13 to OSP14.
* [RFE] Include database validations as part of the upgrade workflow from OSP13 to OSP14.
* [RFE] Document the processes and workflows for the operator to follow when running a major upgrade from OSP13 to OSP14.
Allow testing the Fast Forward Upgrade workflow from OSP13 to OSP14.
* [RFE] Document the processes and workflows for running a Fast Forward Upgrade from OSP13 to OSP14.
Allow operators to execute tasks from OSP14 Undercloud on a OSP13 Overcloud (Mixed versions)
* [RFE] Validate backwards compatibility to execute tasks from a OSP13 Overcloud from a OSP14 Undercloud.
* [RFE] Document the mixed version capabilities.
Allow TripleO developers to test their patches against the update/upgrade workflows.
* [RFE] Have Upstream CI jobs to test minor updates.
* [RFE] Have Upstream CI jobs to test major upgrades.
* [RFE] Have Upstream CI jobs to test Fast Forward Upgrades.
Allow users/QE to try updates and upgrades timely (catch up updates/upgrades with the upstream release cycle)
* How can we get an RFE for this?!?
* Reduce downtime and risk.
* Keep to the scheduled maintenance window and execute:
- Update/Upgrade workflows.
- Rollback if something went wrong.
* After the time window, the environment must be usable.
* Data plane 100% up.
* Have the ability to control the order of reboots in the minor update.
* Provide hooks, after minor updates, to be able to deal with OVS version upgrade problems.
The impact of a failure when running any update/upgrade workflow depends on the execution target:
* Undercloud: The severity of a failure in the Undercloud is low because it can be fixed easily by restoring the node or simply reinstalling it. This can affect operations against the Overcloud, but services and business functions should still work.
* Overcloud: The severity of a failure in the Overcloud when running the update/upgrade workflows is high/critical because it might affect both the control and data plane services, thus, this can affect both services and business functions for the customers.
With RHOSP 14, the goal is to:
* Having the workflows related to the DFG:Upgrades working.
* Having the workflows validated and stabilized.
* Ensure partner integrations with NFV for the updates/upgrades workflows are stable enough to allow them to start their own tests.
* Having proactive fixes and improvements instead of reactive bug fixing due to the lack of upstream CI.