mirror of
https://github.com/hanzo-ml/kuberay.git
synced 2026-10-30 19:48:47 +00:00
* refactor: Standardize main.go style - Add package doc - Move flag declarations into a var() block - Add description strings to each flag - Add section dividers for readability - Group imports into stdlib / third-party / project - Remove dead initialization of runtimeClassConfigPath (overwritten by flag default) No behavior change. Signed-off-by: JiangJiaWei1103 <[email protected]> * feat: Lazy session processing via Supervisor with isDead probe - Add a Supervisor as /enter_cluster broker - Use in-process loaded set to record processed sessions - Add a Pipeline that wraps the parse steps - Remove eager processing on startup and ticker Signed-off-by: JiangJiaWei1103 <[email protected]> * test: Cover Supervisor error propagation and no-retry contracts Signed-off-by: JiangJiaWei1103 <[email protected]> * fix: Surface storage outage error in ProcessSingleSession Signed-off-by: JiangJiaWei1103 <[email protected]> * fix: Handle isDead for RayCluster recreation with the same namespacedname Signed-off-by: JiangJiaWei1103 <[email protected]> * fix: Use live sentinel to avoid querying empty in-mem state Signed-off-by: JiangJiaWei1103 <[email protected]> * refactor: Skip singleflight for already-loaded sessions Signed-off-by: JiangJiaWei1103 <[email protected]> * refactor: Reuse k8s client Signed-off-by: JiangJiaWei1103 <[email protected]> * fix: Surface errors of two subsystems Signed-off-by: JiangJiaWei1103 <[email protected]> * feat: Add SessionSnapshot schema type Signed-off-by: JiangJiaWei1103 <[email protected]> * feat: Add SnapshotLoader (LRU cache) Signed-off-by: JiangJiaWei1103 <[email protected]> * feat: Add GetRawEventsByJobID for snapshot building Signed-off-by: JiangJiaWei1103 <[email protected]> * feat: Pipeline returns SessionSnapshot, Supervisor uses SnapshotLoader Signed-off-by: JiangJiaWei1103 <[email protected]> * refactor: Migrate /tasks family to SnapshotLoader Signed-off-by: JiangJiaWei1103 <[email protected]> * refactor: Migrate /actors family to SnapshotLoader Signed-off-by: JiangJiaWei1103 <[email protected]> * refactor: Migrate /jobs family to SnapshotLoader Signed-off-by: JiangJiaWei1103 <[email protected]> * refactor: Migrate /nodes family to SnapshotLoader Signed-off-by: JiangJiaWei1103 <[email protected]> * refactor: Migrate /events and /api/cluster_status to SnapshotLoader Signed-off-by: JiangJiaWei1103 <[email protected]> * docs: Describe LRU cache layer Signed-off-by: JiangJiaWei1103 <[email protected]> * fix: Add missing copy Signed-off-by: JiangJiaWei1103 <[email protected]> * refactor: Migrate log file resolvers to SnapshotLoader Signed-off-by: JiangJiaWei1103 <[email protected]> * refactor: Clean up Signed-off-by: JiangJiaWei1103 <[email protected]> * refactor: Clean up Signed-off-by: JiangJiaWei1103 <[email protected]> * refactor: Streamline LRU cache impl Signed-off-by: JiangJiaWei1103 <[email protected]> * refactor: Keep dir structure simple Signed-off-by: JiangJiaWei1103 <[email protected]> * refactor: Remove dead codes Signed-off-by: JiangJiaWei1103 <[email protected]> * refactor: Make eventserver focused on event ingestion Signed-off-by: JiangJiaWei1103 <[email protected]> * refactor: Fatal instead of panic Signed-off-by: JiangJiaWei1103 <[email protected]> * refactor: Use clusterSessionKey Signed-off-by: JiangJiaWei1103 <[email protected]> * refactor: Guide users to re-enter the cluster Signed-off-by: JiangJiaWei1103 <[email protected]> * refactor: Avoid over-engineering Signed-off-by: JiangJiaWei1103 <[email protected]> * refactor: Clean up Signed-off-by: JiangJiaWei1103 <[email protected]> * fix: Ensure snapshot first Signed-off-by: JiangJiaWei1103 <[email protected]> * refactor: Simplify log events handling Signed-off-by: JiangJiaWei1103 <[email protected]> * refactor: Better naming and align the orig behavior Signed-off-by: JiangJiaWei1103 <[email protected]> * refactor: Treat snap as SSOT Signed-off-by: JiangJiaWei1103 <[email protected]> * fix: Correct session key example Signed-off-by: JiangJiaWei1103 <[email protected]> * refactor: Align function name Signed-off-by: JiangJiaWei1103 <[email protected]> * fix: Fix test fixture and remove unused tests Signed-off-by: JiangJiaWei1103 <[email protected]> * refactor: Rename cache priming to align with getter Signed-off-by: JiangJiaWei1103 <[email protected]> * fix: Support per-request view of snapshot to avoid race Signed-off-by: JiangJiaWei1103 <[email protected]> * fix: GC in-mem maps Signed-off-by: JiangJiaWei1103 <[email protected]> * refactor: Remove dead EventHandler accessors made unreachable by per-call handler Signed-off-by: JiangJiaWei1103 <[email protected]> * refactor: Align getter deep copy Signed-off-by: JiangJiaWei1103 <[email protected]> * feat: Support absolute ttl expiry in lru cache Signed-off-by: JiangJiaWei1103 <[email protected]> * feat: Support idle TTL Signed-off-by: JiangJiaWei1103 <[email protected]> * docs: Clarify ttl semantics Signed-off-by: JiangJiaWei1103 <[email protected]> * fix: Fix tests Signed-off-by: JiangJiaWei1103 <[email protected]> * refactor: Remove unused field Signed-off-by: JiangJiaWei1103 <[email protected]> --------- Signed-off-by: JiangJiaWei1103 <[email protected]>