Stack lifecycle scheduler hint blueprint¶
A heat provider may have a need for custom code to examine stack requests prior to performing the operations to create or update a stack. After the custom code completes, the provider may want to provide hints to the nova scheduler with stack related identifiers, for processing by any custom scheduler plug-ins configured for nova.
A heat provider may have a need for custom code to examine stack requests prior to performing the operations to create or update a stack. An example would be a holistic scheduler that schedules a stack’s member compute resources as group. This would be done using a custom plugin invoked through the stack lifecycle plugpoint. After the custom code completes, when the create or update is being processed, any custom schedulers configured for nova would need to map nova create requests back to any decisions made during the call to the custom stack lifecycle plugin. Current heat includes no identifiers in a nova create request that can be used to map back to a Server or Instance resource within a heat stack.
It is out of scope for this spec, but worth noting that cinder scheduler hints are now supported by heat and may need similar treatment. See https://review.openstack.org/#/c/126282/ and https://review.openstack.org/#/c/126298/
When heat processes a stack, and the feature is enabled, the stack id, root stack id, stack resource id, stack resource name and the path in the stack (as a list of tuples, (stackresourcename, stackname)) will be passed to nova by heat as scheduler hints, to the configured schedulers for nova.
The behavior changes will be optional, default disabled, and controlled through a new heat config variable.
These five scheduler hints will be added to server creates done using either resource class Server (OS::Nova::Server) or resource class Instance (AWS::EC2::Instance). heat_root_stack_id will be set to the id of the root stack of the resource, heat_stack_id will be set to the id of the resource’s parent stack, heat_stack_name will be set to the id of the resource’s parent stack, heat_path_in_stack will be set to a list of tuples, (stackresourcename, stackname) with list being (None, rootstackname), and heat_resource_name will be set to the resource’s name
No reasonable alternatives were identified. Similar function could be achieved if the lifecycle plugin modified the stack (and changes were persisted). This would be bad behavior. It would conflict with convergence when it lands, and scheduler decisions would become visible to the heat user (unless somehow redacted on query).
A patch comprising a full implementation of the blueprint (https://review.openstack.org/#/c/96889/) is already being reviewed, under the old pre-spec process.
- Primary assignee:
William C. Arnold (barnold-8)
- Target Milestone for completion:
Implementation: https://review.openstack.org/#/c/96889/ Documentation: Add good documentation to heat in tree docs