Merge remote-tracking branch 'origin/fix/db-off-data-plane-469' into fix/db-off-data-plane-469
tracker-policy-pr / check-pr (pull_request) Successful in 9s
test / integration-docker (pull_request) Successful in 16s
test / unit (pull_request) Successful in 43s
lint / lint (push) Successful in 2m44s
test / integration-firecracker (pull_request) Successful in 3m42s
test / coverage (pull_request) Successful in 15s
test / publish-infra (pull_request) Has been skipped

# Conflicts:
#	tests/unit/test_macos_infra.py
This commit is contained in:
2026-07-25 18:01:49 -04:00
2 changed files with 29 additions and 12 deletions
+17 -5
View File
@@ -168,13 +168,19 @@ def stop() -> None:
recorded VMMs AND any orphaned firecracker still bound to either infra
config — the PID files drift after crashes / out-of-band kills, and a
survivor would hold a link's TAP so the next boot dies with "tap …
Resource busy". Drops the version marker so a stopped pair is never
treated as adoptable."""
Resource busy". Also reaps a surviving *legacy* single combined-VM (the
pre-split layout booted from `<infra>/config.json` on the orchestrator
link): the two-VM `stop` otherwise wouldn't know about it, and it would
hold the orchestrator TAP so the first split boot fails — so the cutover
is self-healing, no manual host teardown. Drops the version marker so a
stopped pair is never treated as adoptable."""
_kill_pidfile(_orch_dir())
_kill_pidfile(_gw_dir())
_kill_pidfile(_infra_dir()) # legacy pre-split combined VM (migration)
_kill_infra_firecrackers()
_pid_file(_orch_dir()).unlink(missing_ok=True)
_pid_file(_gw_dir()).unlink(missing_ok=True)
_pid_file(_infra_dir()).unlink(missing_ok=True) # legacy
_version_file().unlink(missing_ok=True)
@@ -357,9 +363,15 @@ def _kill_infra_firecrackers(proc_root: Path = Path("/proc")) -> None:
"""SIGKILL any firecracker VMM whose `--config-file` is one of this host's
infra configs (orchestrator or gateway), independent of the PID files —
reaps orphans it lost track of so both links' TAPs are free to rebind.
Scoped to the two infra config paths, so the interactive pool's agent VMs
(other config paths) are untouched."""
cfgs = {str(_orch_dir() / "config.json"), str(_gw_dir() / "config.json")}
Also matches the *legacy* pre-split combined-VM config (`<infra>/config.json`)
so a surviving old singleton is cleared off the orchestrator link during the
cutover. Scoped to these infra config paths, so the pool's agent VMs (other
config paths) are untouched."""
cfgs = {
str(_orch_dir() / "config.json"),
str(_gw_dir() / "config.json"),
str(_infra_dir() / "config.json"), # legacy pre-split combined VM
}
for entry in proc_root.iterdir():
if not entry.name.isdigit():
continue