PR #25263: fix(gateway): restart via service in containers

• Repository: NousResearch/hermes-agent
• Author: he-yufeng
• State: closed | merged: False
• Link: https://github.com/NousResearch/hermes-agent/pull/25263

Description (problem / solution / changelog)

Summary

Closes #25217.

/restart currently uses a detached helper whenever the gateway is not running under systemd. That works for a normal shell process, but not for Docker/Podman: once the gateway exits as PID 1, the container stops and the detached helper goes with it.

This routes container restarts through the existing service-restart path instead:

• systemd: unchanged, detached=False, via_service=True
• Docker/Podman: now also detached=False, via_service=True
• plain non-container process: still uses the detached helper

The service path exits with the existing restart code, so a Docker/Podman restart policy such as unless-stopped or on-failure can bring the container back.

Tests

python -m py_compile gateway\run.py tests\gateway\test_restart_notification.py tests\gateway\test_restart_drain.py
python -m ruff check gateway\run.py tests\gateway\test_restart_notification.py tests\gateway\test_restart_drain.py
python -m pytest -o addopts='' tests\gateway\test_restart_notification.py tests\gateway\test_restart_drain.py::test_restart_command_while_busy_requests_drain_without_interrupt -q --basetemp .tmp\pytest -p no:cacheprovider

Result: 27 passed.

Note: the full tests\gateway\test_restart_drain.py file has an existing Windows-only failure in test_launch_detached_restart_command_uses_setsid: this Windows run enters the sys.executable -c detached watcher branch, while that test expects the POSIX setsid bash path.

Changed files

• gateway/run.py (modified, +7/-6)
• tests/gateway/test_restart_drain.py (modified, +1/-0)
• tests/gateway/test_restart_notification.py (modified, +24/-0)

PR #25264: fix(gateway): detect Docker/Podman containers for /restart service path

• Repository: NousResearch/hermes-agent
• Author: zccyman
• State: open | merged: False
• Link: https://github.com/NousResearch/hermes-agent/pull/25264

Description (problem / solution / changelog)

Summary

The /restart gateway command only detected systemd (INVOCATION_ID) when choosing between the detached-subprocess and service-restart approaches. In Docker/Podman containers, the detached subprocess approach fails because the container stops when PID 1 exits, killing the helper before it can restart anything.

Problem

In a Docker container (e.g. nousresearch/hermes-agent:latest on Unraid):

1. User sends /restart
2. Gateway exits to spawn a detached restart helper
3. tini (PID 1) exits because the gateway exited
4. Docker stops the container
5. The detached helper is killed before it can restart anything
6. Gateway stays dead until manually restarted

Fix

Detect Docker (/.dockerenv) and Podman (/run/.containerenv) alongside systemd. In container environments, use the service restart path (exit code 75) instead of the detached subprocess approach. Combined with a restart policy (unless-stopped or on-failure), Docker/Podman handles the restart correctly.

_under_service = bool(
    os.environ.get("INVOCATION_ID")       # systemd
    or os.path.isfile("/.dockerenv")       # Docker
    or os.path.isfile("/run/.containerenv") # Podman
)

Testing

• All 132 gateway restart-related tests pass

Files Changed

Closes #25217

Changed files

• gateway/run.py (modified, +11/-1)