headroom¶

The headroom processor provides cluster headroom post-processing analytic parity for Turbonomic’s native headroom calculations, while enabling extended headroom customization through custom template mappings.

Enumerations¶

class vmtplanner.processors.headroom.HeadroomMode(value)¶

Headroom Calculation Modes

AVERAGE = 2¶: Split evenly amongst all templates in a group, i.e. average all templates

SEPARATE = 1¶: Per-template headroom

SUM = 3¶: Combined templates in a group, i.e. summed

Classes¶

class vmtplanner.processors.headroom.HeadroomEncoder(*, skipkeys=False, ensure_ascii=True, check_circular=True, allow_nan=True, sort_keys=False, indent=None, separators=None, default=None)¶

Headroom results encoder for JSON output

Example

with open(OUTFILE, 'w') as fp:
    json.dump(plan.headroom(), fp, indent=2, cls=HeadroomEncoder)

class vmtplanner.processors.headroom.Group(name=None, uuid=None)¶

Headroom group

Groups provide partitioning within a cluster, and are not required if no subdivision of the cluster is necessary, as each cluster has a default group for all ungrouped members.. Groups need not be created per cluster either, as all entities will be partitioned on cluster boundaries before being assigned to their respective groups. Groups cannot be used to create super clusters.

Parameters

name (str) – Group name. One of name or uuid is required.
uuid (str) – Group UUID.

name¶

Group display name.

Type: str

uuid¶

Group UUID.

Type: str

type¶

Trubonomic “groupType” of the group:

Type: str

members¶

List of group real-time market members.

Type: list

Raises: ValueError – If both name and uuid are None.

Notes

One of name or uuid is required to lookup the group, if both are provided uuid will override.

class vmtplanner.processors.headroom.Template(name=None, uuid=None, targets=None, clusters=None)¶

Headroom template

Templates may be linked to both groups and clusters, and will apply to all members at the level applied respecitvely. Because groups can span across clusters, you may want to assign specific templates to specific cluster group combinations. You can specify a cluster or clusters to lock a template to, which will cause the template to only be applied to group members of the listed clusters.

Parameters

name (str) – Template name. One of name or uuid is required.
uuid (str) – Template UUID.
targets (list) – List of groups and/or clusters to apply the template to.
clusters (list, optional) – List of clusters to limit the template to.

name¶

Template display name.

Type: str

uuid¶

Template UUID.

Type: str

targets¶

List of group names or UUIDs to apply the template to.

Type: list

clusters¶

List of clusters to limit the template to.

Type: list

cpu¶

Template CPU value.

Type: Decimal

cpu_provisioned¶

Template CPU provisioned value.

Type: Decimal

mem¶

Template memory value.

Type: Decimal

mem_provisioned¶

Template memory provisioned value.

Type: Decimal

storage_amount¶

Template storage value.

Type: Decimal

storage_provisioned¶

Template storage provisioned value.

Type: Decimal

Raises

TypeError – If the retrieved template is not a VirtualMachine template.
ValueError – If both name and uuid are None; or no targets provided.

Notes

One of name or uuid is required to lookup the template, if both are provided uuid will override.

class vmtplanner.processors.headroom.Cluster(connection, uuid, name, members=None, realtime_members=None, mode=<HeadroomMode.SEPARATE: 1>)¶

Headroom cluster

Individual cluster objects, used by a ClusterHeadroom plan.

Parameters

connection (Connection) – Connection or Session.
uuid (int) – Cluster UUID.
name (str) – Cluster display name.
members (list, optional) – List of cluster member UUIDs.
realtime_members (list, optional) – List of realtime market member UUIDs.
mode (HeadroomMode, optional) – Headroom calculation mode. (default: HeadroomMode.SEPARATE)

name¶

Cluser display name.

Type: str

datacenter¶

Datacenter the cluster belongs to.

Type: str

groups¶

Dictionary of cluster groups for headroom analysis.

Type: dict

growth¶

Cluster growth.

Type: float

members¶

List of cluster member UUIDs.

Type: list

mode¶

Headroom calculation mode.

Type: HeadroomMode

realtime_members¶

List of realtime market member UUIDs.

Type: list

uuid¶

Cluster UUID.

Type: str

class vmtplanner.processors.headroom.ClusterHeadroom(connection, spec=None, market='Market', scope=None, groups=None, templates=None, growth_lookback=7, mode=<HeadroomMode.SEPARATE: 1>)¶

Cluster headroom plan

In basic form, this provides cluster headroom parity with Turbonomics native 10 Day average templates. When combined with groups and templates, ClusterHeadroom provides highly customizable headroom analysis.

Parameters

connection (Connection) – Connection or Session.
spec (PlanSpec, optional) – Settings override to apply to the market. Default behavior is to run a balance plan first.
market (str, optional) – Base market UUID to apply the settings to. (default: Market)
scope (list, optional) – Scope of the plan market. If None, then a list of all clusters in the given market will be used.
groups (list) – List of Group objects.
templates (list) – List of :py:class`Template` objects.
growth_lookback (int) – Number of days to use for growth calcuation.
mode (HeadroomMode, optional) – Headroom calculation mode.

commodities¶

Commodities to calculate for headroom.

Type: list

clusters¶

List of clusters in the plan.

Type: list

groups¶

List of Group objects.

Type: list

growth_lookback¶

VM growth period in days.

Type: int

mode¶

Headroom calculation mode.

Type: HeadroomMode

templates¶

List of :py:class`Template` objects.

Type: list

tempalte_commodity¶

Dictionary map of commodities to template attributes.

Type: dict