rlightning.utils.placement.placement_strategies

Placement strategies for component scheduling.

This module provides different strategies for placing components on cluster resources: - DefaultPlacementStrategy: No specific placement, uses Ray’s default scheduling - ResourcePoolPlacementStrategy: Base class for resource pool-based placement - DisaggregatePlacementStrategy: Separate pools for Train+Buffer and Eval+Env - ColocatedPlacementStrategy: Shared pool for all components

class rlightning.utils.placement.placement_strategies.ColocatedPlacementStrategy(scheduling: ComponentScheduling)

Bases: ResourcePoolPlacementStrategy

Colocated placement strategy.

All components share a global resource pool. Workers are distributed across nodes with consideration for resource utilization.

create_placement_groups(resource_pools: Dict[str, Dict[str, Any]] | None = None, max_colocate_count: int = 10, **kwargs) → Dict[str, MockModule('ray.util.placement_group.PlacementGroup')]

Create placement groups for colocated components.

Parameters:

• resource_pools – Dictionary with “global_pool” configuration.

• max_colocate_count – Maximum number of components per node.

• **kwargs – Additional strategy options kept for interface compatibility.

class rlightning.utils.placement.placement_strategies.DefaultPlacementStrategy(scheduling: ComponentScheduling)

Bases: PlacementStrategy

Default placement strategy with no specific placement groups.

Uses node affinity for buffer workers when multiple are needed, but otherwise relies on Ray’s default scheduling.

buffer_strategys

List of scheduling strategies for buffer workers.

_node_info

Node information for resource-aware choices.

create_placement_groups() → Dict[str, MockModule('ray.util.placement_group.PlacementGroup')]

Create placement groups based on the strategy.

For multiple buffer storages, creates node affinity strategies to place each buffer on a different node.

Returns:

Empty dictionary (no placement groups created).

get_scheduling_strategy(component_type: str, worker_index: int = 0) → Any

Get scheduling strategy for a component.

Parameters:

• component_type – Type of component.

• worker_index – Index of the worker.

Returns:

Node affinity strategy for buffers, ‘DEFAULT’ otherwise.

class rlightning.utils.placement.placement_strategies.DisaggregatePlacementStrategy(scheduling: ComponentScheduling)

Bases: ResourcePoolPlacementStrategy

Disaggregate placement strategy.

Allocates separate resource pools for: - train_pool: Train workers + Buffer workers (colocated) - rollout_pool: Eval workers + Env workers (colocated)

This provides resource isolation between training and evaluation.

create_placement_groups(resource_pools: Dict[str, Dict[str, Any]] | None = None, **kwargs) → Dict[str, MockModule('ray.util.placement_group.PlacementGroup')]

Create placement groups based on resource pools.

Parameters:

• resource_pools – Dictionary with “train_pool” and “rollout_pool” configurations.

• **kwargs – Additional strategy options kept for interface compatibility.

class rlightning.utils.placement.placement_strategies.PlacementStrategy(scheduling: ComponentScheduling)

Bases: ABC

Abstract base class for placement strategies.

Defines the interface for creating Ray placement groups and determining scheduling strategies for different component types.

scheduling

Cluster scheduling configuration.

placement_groups

Created placement groups by name.

storage_to_train_workers

Mapping from storage index to train worker indices.

cleanup() → None

Clean up placement groups.

abstractmethod create_placement_groups() → Dict[str, MockModule('ray.util.placement_group.PlacementGroup')]

Create placement groups based on the strategy.

Returns:

Dictionary mapping group names to PlacementGroup instances.

abstractmethod get_scheduling_strategy(component_type: str, worker_index: int = 0) → Any

Get scheduling strategy for a specific component.

Parameters:

• component_type – Type of component (‘env’, ‘train’, ‘eval’, ‘buffer’).

• worker_index – Index of the worker within its type.

Returns:

Scheduling strategy (PlacementGroupSchedulingStrategy or ‘DEFAULT’).

get_storage_to_train_workers() → Dict[int, List[int]]

Return storage -> train worker mapping (may be empty if not used).

Returns:

Dictionary mapping storage indices to lists of train worker indices.

class rlightning.utils.placement.placement_strategies.ResourcePoolPlacementStrategy(scheduling: ComponentScheduling)

Bases: PlacementStrategy

Base class for resource pool-based placement strategies.

Provides common functionality for creating placement groups from resource pools by reading component_types from pool config and scheduling info from self.scheduling.

create_placement_groups(resource_pools: Dict[str, Dict[str, Any]] | None = None, **kwargs) → Dict[str, MockModule('ray.util.placement_group.PlacementGroup')]

Create placement groups based on the strategy.

get_node_component_distribution() → Dict[str, Dict[str, Dict[str, Any]]]

Return the planned component distribution per node.

get_scheduling_strategy(component_type: str, worker_index: int = 0) → Any

Get scheduling strategy for a component.