All issues

Search results

Firecracker runs at Debug, so a vsock-heavy guest writes gigabytes an hour into firecracker.log

Status: Open.

#969 In ejc3/fcvm;

· ejc3 opened on Sep 20, 2026

podman prepare checks snapshot disk space only after the whole boot

Status: Open.

#968 In ejc3/fcvm;

· ejc3 opened on Sep 20, 2026

Health check: an https:// URL is probed as plain HTTP on port 443

Status: Open.

#958 In ejc3/fcvm;

· ejc3 opened on Sep 20, 2026

Build Kernels follow-ups from the review of #952

Status: Open.

#957 In ejc3/fcvm;

· ejc3 opened on Sep 20, 2026

Working-set replay at large-guest scale: single thread, cold page cache, empty first clone, growing set

Status: Open.

#955 In ejc3/fcvm;

· ejc3 opened on Sep 19, 2026

Snapshot key includes the health-check URL, so prepare and run cannot share the post-import snapshot

Status: Open.

#953 In ejc3/fcvm;

· ejc3 opened on Sep 19, 2026

Restore acknowledgement waits on serial socket destroys and three ip forks

Status: Open.

#954 In ejc3/fcvm;

· ejc3 opened on Sep 19, 2026

Move the guest kernel past 6.18.50: 6.18.52 carries the VNCR fix, so the carried patch and its fixture have to go

Status: Open.

#941 In ejc3/fcvm;

· ejc3 opened on Sep 18, 2026

Recover corpus replay cleanup after root supervisor SIGKILL

Status: Open.

#895 In ejc3/fcvm;

· ejc3 opened on Aug 31, 2026

Runner autoscaler kills runners mid-job at the 12h age cap, and the failure reads as an AWS spot reclaim

Status: Open.

#884 In ejc3/fcvm;

· ejc3 opened on Aug 29, 2026

arm64 runner wedged: firecracker D-state in synchronize_srcu (mmu_notifier_unregister/kvm_destroy_vm) after the warm SnapshotEnabled pass

Status: Open.

#871 In ejc3/fcvm;

· ejc3 opened on Aug 28, 2026

Fuzz oracles for companion-process death and hostile host config

Status: Open.

#852 In ejc3/fcvm;

· ejc3 opened on Aug 16, 2026