Skip to content

fix: spot interruption detection #2391

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
wants to merge 2 commits into
base: master
Choose a base branch
from
Open

fix: spot interruption detection #2391

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
c81c11a
fix for sending SIGTERM to main process
saikonen Apr 25, 2025
fd872fd
use SIGUSR1 instead of SIGTERM for signaling about receiving the spot…
saikonen May 2, 2025
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
2 changes: 2 additions & 0 deletions metaflow/plugins/aws/batch/batch_decorator.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -299,6 +299,8 @@ def task_pre_step(
self._save_logs_sidecar.start()

# Start spot termination monitor sidecar.
# TODO: A nicer way to pass the main process id to a Sidecar, in order to allow sidecars to send signals back to the main process.
os.environ["MF_MAIN_PID"] = str(os.getpid())
current._update_env(
{"spot_termination_notice": "/tmp/spot_termination_notice"}
)
Expand Down
2 changes: 2 additions & 0 deletions metaflow/plugins/kubernetes/kubernetes_decorator.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -559,6 +559,8 @@ def task_pre_step(
self._save_logs_sidecar.start()

# Start spot termination monitor sidecar.
# TODO: A nicer way to pass the main process id to a Sidecar, in order to allow sidecars to send signals back to the main process.
os.environ["MF_MAIN_PID"] = str(os.getpid())
current._update_env(
{"spot_termination_notice": "/tmp/spot_termination_notice"}
)
Expand Down
5 changes: 4 additions & 1 deletion metaflow/plugins/kubernetes/spot_monitor_sidecar.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -21,6 +21,9 @@ def __init__(self):
self._token = None
self._token_expiry = 0

# Due to nesting, os.getppid is not reliable for fetching the main task pid
self.main_pid = int(os.getenv("MF_MAIN_PID", os.getppid()))

if self._is_aws_spot_instance():
self._process = Process(target=self._monitor_loop)
self._process.start()
Expand Down Expand Up @@ -71,7 +74,7 @@ def _monitor_loop(self):
if response.status_code == 200:
termination_time = response.text
self._emit_termination_metadata(termination_time)
os.kill(os.getppid(), signal.SIGTERM)
os.kill(self.main_pid, signal.SIGUSR1)
break
except (requests.exceptions.RequestException, requests.exceptions.Timeout):
pass
Expand Down
Loading