Skip to content

fix(spurctld): reclaim agent allocations the controller no longer tracks on that node #623

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
yansun1996 wants to merge 3 commits into
base: main
Choose a base branch
from
+293 −27
Open

fix(spurctld): reclaim agent allocations the controller no longer tracks on that node #623

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
99c9914
fix(spurctld): reclaim agent allocations for jobs aged out of the job…
yansun1996 Aug 12, 2026
635a7c6
fix(spurctld): let a node release a job the controller runs elsewhere
yansun1996 Aug 12, 2026
d4c59a6
fix(spurctld): treat an active job's empty nodelist as non-authoritative
yansun1996 Aug 13, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
9 changes: 6 additions & 3 deletions crates/spurctld/src/cluster.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -892,9 +892,10 @@ impl ClusterManager {
self.jobs.read().get(&job_id).cloned()
}

/// A job's state by ID, without cloning the whole `Job`.
pub fn job_state(&self, job_id: JobId) -> Option<JobState> {
self.jobs.read().get(&job_id).map(|j| j.state)
/// Next id this controller would assign. Ids at or above it were never
/// issued here; ids below it were not necessarily issued either.
pub fn peek_next_job_id(&self) -> JobId {
self.next_job_id.load(Ordering::Relaxed)
}

/// Get a job by ID, synthesizing an aggregate record for an array *parent*
Expand Down Expand Up @@ -5528,6 +5529,8 @@ impl StateMachineApply for ClusterManager {
// (NOT config-derived like license_pool/burst_buffer) — restore them.
*self.k0s.write() = snap.k0s;

// Must stay inside the `jobs` write guard: a reader seeing the cleared
// map with this watermark would treat every live job as reclaimable.
self.next_job_id.store(next_id, Ordering::Relaxed);

// Re-evaluate partition membership and NodeConfig policy
Expand Down
Loading