mirror of
https://github.com/CloudNebulaProject/reddwarf.git
synced 2026-04-10 13:20:40 +00:00
Image management: - ImageCatalog backed by KVStore with register/resolve/list/delete - CLI `image import`, `image list`, `image delete` subcommands - PodController resolves container image field to local path (tarball → lx_image_path, ZFS snapshot → clone_from) Scheduler runtime metrics: - compute_allocated_resources() sums requests of all scheduled pods per node - PodFitsResources filter subtracts used resources from allocatable - LeastAllocated/BalancedAllocation scorers account for existing load - Pod count limits enforced against node max-pods - Allocated resources updated within scheduling cycle for multi-pod batches https://claude.ai/code/session_016QLFjAyYGzMPbBjEGMe75j
18 lines
514 B
Rust
18 lines
514 B
Rust
//! Reddwarf Scheduler - Pod to Node scheduling
|
|
//!
|
|
//! This crate provides:
|
|
//! - Pod scheduling algorithm
|
|
//! - Filter predicates (resource requirements, node selectors)
|
|
//! - Scoring functions (least allocated)
|
|
//! - Pod binding to nodes
|
|
|
|
pub mod error;
|
|
pub mod filter;
|
|
pub mod scheduler;
|
|
pub mod score;
|
|
pub mod types;
|
|
|
|
// Re-export commonly used types
|
|
pub use error::{Result, SchedulerError};
|
|
pub use scheduler::Scheduler;
|
|
pub use types::{FilterResult, NodeAllocatedResources, SchedulingContext, ScoreResult};
|