Difference between revisions of "ExtensibleResourceTracking"

Revision as of 21:36, 28 November 2013

Summary - (in progress)

There are a number of defined resources in Nova, including disk, memory and number of vCPU. The purpose of this specification is to support an extensible set of resources so that new resources can be defined for compute nodes and resource requirements for instances. These will be made available to the scheduler and allocated by the compute manager.

The resources of a compute node have limited capacity. Most, including disk, memory and vCPU, map to a physical counterpart, typically via a virtualisation technology, but it is also possible to create entirely abstract resource types that do not. As an example, a provider may chose that only one instances of a certain type should run on any compute node. This trivial example can be supported by defining a new resource for compute nodes and associating a requirement for that resource with the instance type.

Each instance has resource requirements specified as the quantity of each type of resource it needs. These are used by the scheduler to match it to compute nodes with sufficient available resources.

The compute manager on a host is aware of the resources available at the compute node and is responsible for their allocation and management. Because it has definitive knowledge of what resources are actually in use, the compute manager has the final say as to whether a compute node is able to host an instance.

We will need:

a model of a resource
a model of a resource requirement
a naming scheme to allow resource requirements to be related to resources
changes for compute nodes
- resource configuration - to define extended resources
- resource_tracker - to track extended resources
- claims - to test requirements against available extended resources
changes for instances
- resource requirements in flavors
- resource requirements in requests
- resource requirements in instances
changes for the database
- extended resources at compute_nodes
- extended resource requirements in instances
changes for scheduler (filter scheduler)
- host manager - to access extended resources
- filters - examples using extended resources
- weighers - examples using extended resources

Resource Models and Naming

Resource Model

A resource has the following attributes:

name - used to identify this resource
type - a resource type name for human viewing
units - a resource unit name for human viewing
description - a short description for human viewing

Each resource at a compute node is represented by its total amount and the amount used - this is recorded by the resource tracker and used in scheduling and allocation decisions.

resource total = value (configured or discovered - total amount of resource)
resource used = value (tracked - amount of resource currently used)

Each resource required for a particular instance is represented by a single value.

resource required = value (configured - amount required)

The two final quantities associated with resources are the commit ratio and the limit. The commit ratio defines how much of a resource can be allocated (a provider may chose to over commit a resource or leave some headroom) and a limit (the total amount that can be committed.

resource commit ratio = value (configured - a proportion of the available resource that can be allocated)
resource limit = value (calculated - the total amount of resource that can be allocated)

@@ Line 29: / Line 29: @@
 === Resource Models and Naming ===
+==== Resource Model ====
 A resource has the following attributes:
 * name - used to identify this resource
@@ Line 36: / Line 38: @@
-The resource_tracker maintains two attributes for each resource of a compute node:
+Each resource at a compute node is represented by its total amount and the amount used - this is recorded by the resource tracker and used in scheduling and allocation decisions.
-* resource total = value (total amount of resource)
+* resource total = value (configured or discovered - total amount of resource)
-* resource used = value (amount of resource currently used)
+* resource used = value (tracked - amount of resource currently used)
-A flavor has a single attribute for each resource it requires:
-* resource required = value (amount required)
-An instance has a single attribute for each resource it requires:
-* resource required = value (from flavor)
-The information known to the resource_tracker is communicated to the scheduler's host_manager and accessed through the host_state:
-* resource total = value (from resource_tracker)
-* resource used = value (from resource_tracker)
-When the scheduler matches an instance create request to a compute node, it uses a resource allocation policy that includes knowledge of how much it can over-commit the resource. It may also chose to ignore resource limits (for example, when forcing an instance to a host). It calculates a resource limit value to guide its decision.
+Each resource required for a particular instance is represented by a single value.
-* resource total = value (from host_state)
+* resource required = value (configured - amount required)
-* resource used = value (from host_state)
-* resource commit ratio = value (configured)
-* resource limit = value (calculated)
-Instance create requests communicate details of the instance and its requirements, including the limit used to calculate if it fits on the host according to the resource allocation policy.
+The two final quantities associated with resources are the commit ratio and the limit. The commit ratio defines how much of a resource can be allocated (a provider may chose to over commit a resource or leave some headroom) and a limit (the total amount that can be committed.
-* resource required = value (from flavor)
+* resource commit ratio = value (configured - a proportion of the available resource that can be allocated)
-* resource limit = value (from scheduler)
+* resource limit = value (calculated - the total amount of resource that can be allocated)