feat(BA-3103): Add ScalingGroup strawberry-base GQL type
#6992
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
jopemachine
changed the title
github-actions
bot
added
size:L
jopemachine
changed the title
ScalingGroup strawberry-base GQL type
jopemachine
force-pushed
the
feat/BA-3103
branch
from
3c34c50 to
04e4427
Compare
Co-authored-by: octodog <mu001@lablup.com>
1 task
jopemachine
force-pushed
the
feat/BA-3103
branch
from
d50a741 to
e87f056
Compare
HyeockJinKim
requested changes
Nov 28, 2025
Comment on lines
+36
to
+37
| @strawberry.field(description="Added in 25.18.0. List scaling groups") | ||
| async def scaling_groups( |
Collaborator
There was a problem hiding this comment.
Is this scaling_groups to be used as a whole? Does it overlap with the existing query?
Comment on lines
+86
to
+107
| description="Session types that can be scheduled in this scaling group. Valid values: 'interactive' , 'batch', 'inference'. Requests for unlisted types are rejected." | ||
| ) | ||
| pending_timeout: float = strawberry.field( | ||
| description="Maximum time in seconds a session can wait in PENDING state before automatic cancellation. Zero means no timeout. Used to prevent indefinite resource waiting when no agents are available." | ||
| ) | ||
| config: JSON = strawberry.field( | ||
| description="Scheduler-specific configuration options. Contents depend on the scheduler implementation (fifo/lifo/drf). Used for advanced scheduling behavior customization." | ||
| ) | ||
| agent_selection_strategy: str = strawberry.field( | ||
| description="Algorithm for selecting target agents when scheduling sessions. 'dispersed' spreads sessions across available agents, 'concentrated' packs sessions onto fewer agents, 'roundrobin' cycles through agents sequentially." | ||
| ) | ||
| agent_selector_config: JSON = strawberry.field( | ||
| description="Configuration for the agent selection strategy. Structure varies by strategy - for example, concentrated strategy may specify endpoint spreading rules." | ||
| ) | ||
| enforce_spreading_endpoint_replica: bool = strawberry.field( | ||
| description="When true with concentrated strategy, forces inference service replicas to be distributed across different agents instead of co-locating on same agent. Improves fault tolerance for model serving." | ||
| ) | ||
| allow_fractional_resource_fragmentation: bool = strawberry.field( | ||
| description="Whether agents accept session requests that allocate fractional resources (e.g., 0.5 GPU) causing resource fragmentation. When false, agents reject sessions that would prevent future efficient resource allocation." | ||
| ) | ||
| route_cleanup_target_statuses: list[str] = strawberry.field( | ||
| description="List of route health statuses that trigger automatic cleanup of service routes. Valid values: 'healthy', 'unhealthy', 'degraded'. Default: ['unhealthy']." |
Collaborator
There was a problem hiding this comment.
Let's use multi line + textwrap for the description.
Comment on lines
+101
to
+102
| description="When true with concentrated strategy, forces inference service replicas to be distributed across different agents instead of co-locating on same agent. Improves fault tolerance for model serving." | ||
| ) |
Collaborator
There was a problem hiding this comment.
It seems good to first explain the description of the value, and then explain what behavior occurs when it is set to true.
Comment on lines
+150
to
+181
| status=ScalingGroupStatus( | ||
| is_active=data.status.is_active, | ||
| is_public=data.status.is_public, | ||
| ), | ||
| metadata=ScalingGroupMetadata( | ||
| description=data.metadata.description, | ||
| created_at=data.metadata.created_at, | ||
| ), | ||
| wsproxy=ScalingGroupNetworkConfig( | ||
| wsproxy_addr=data.wsproxy.wsproxy_addr, | ||
| wsproxy_api_token=data.wsproxy.wsproxy_api_token, | ||
| use_host_network=data.wsproxy.use_host_network, | ||
| ), | ||
| driver=ScalingGroupDriverConfig( | ||
| name=data.driver.name, | ||
| options=data.driver.options, | ||
| ), | ||
| scheduler=ScalingGroupSchedulerConfig( | ||
| name=data.scheduler.name, | ||
| options=ScalingGroupSchedulerOptions( | ||
| allowed_session_types=[ | ||
| st.value for st in data.scheduler.options.allowed_session_types | ||
| ], | ||
| pending_timeout=data.scheduler.options.pending_timeout.total_seconds(), | ||
| config=data.scheduler.options.config, | ||
| agent_selection_strategy=data.scheduler.options.agent_selection_strategy.value, | ||
| agent_selector_config=data.scheduler.options.agent_selector_config, | ||
| enforce_spreading_endpoint_replica=data.scheduler.options.enforce_spreading_endpoint_replica, | ||
| allow_fractional_resource_fragmentation=data.scheduler.options.allow_fractional_resource_fragmentation, | ||
| route_cleanup_target_statuses=data.scheduler.options.route_cleanup_target_statuses, | ||
| ), | ||
| ), |
Collaborator
There was a problem hiding this comment.
It seems good to make from_dataclass for each type.
Comment on lines
+191
to
+194
| DESCRIPTION = "description" | ||
| CREATED_AT = "created_at" | ||
| IS_ACTIVE = "is_active" | ||
| IS_PUBLIC = "is_public" |
Collaborator
There was a problem hiding this comment.
Description order doesn't seem strictly necessary.
Comment on lines
+98
to
+103
| @staticmethod | ||
| def description(ascending: bool = True) -> QueryOrder: | ||
| if ascending: | ||
| return ScalingGroupRow.description.asc() | ||
| else: | ||
| return ScalingGroupRow.description.desc() |
Collaborator
There was a problem hiding this comment.
It seems good that the order of the description is not provided from the backend.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
resolves #6867 (BA-3103)
Checklist: (if applicable)
📚 Documentation preview 📚: https://sorna--6992.org.readthedocs.build/en/6992/
📚 Documentation preview 📚: https://sorna-ko--6992.org.readthedocs.build/ko/6992/