🐛 fixes issue with agent not executing in correct container (#8256)

Co-authored-by: Andrei Neagu <[email protected]>

Author: GitHK

services/agent/src/simcore_service_agent/services/backup.py

@@ -1,6 +1,7 @@
 import asyncio
+import json
 import logging
-import os
+import socket
 import tempfile
 from asyncio.streams import StreamReader
 from datetime import timedelta
@@ -9,6 +10,7 @@
 from typing import Final
 from uuid import uuid4
 
+import httpx
 from fastapi import FastAPI
 from servicelib.container_utils import run_command_in_container
 from settings_library.utils_r_clone import resolve_provider
@@ -112,8 +114,28 @@ def _log_expected_operation(
     _logger.log(log_level, formatted_message)
 
 
+def _get_self_container_ip() -> str:
+    return socket.gethostbyname(socket.gethostname())
+
+
+async def _get_self_container() -> str:
+    ip = _get_self_container_ip()
+
+    async with httpx.AsyncClient(
+        transport=httpx.AsyncHTTPTransport(uds="/var/run/docker.sock")
+    ) as client:
+        response = await client.get("http://localhost/containers/json")
+        for entry in response.json():
+            if ip in json.dumps(entry):
+                container_id: str = entry["Id"]
+                return container_id
+
+    msg = "Could not determine self container ID"
+    raise RuntimeError(msg)
+
+
 async def _ensure_permissions_on_source_dir(source_dir: Path) -> None:
-    self_container = os.environ["HOSTNAME"]
+    self_container = await _get_self_container()
     await run_command_in_container(
         self_container,
         command=f"chmod -R o+rX '{source_dir}'",

services/agent/tests/unit/test_services_backup.py

@@ -15,6 +15,8 @@
 from models_library.projects_nodes_io import NodeID
 from models_library.services_types import ServiceRunID
 from pydantic import NonNegativeInt
+from pytest_mock import MockerFixture
+from servicelib.container_utils import run_command_in_container
 from simcore_service_agent.core.settings import ApplicationSettings
 from simcore_service_agent.services.backup import backup_volume
 from simcore_service_agent.services.docker_utils import get_volume_details
@@ -42,7 +44,7 @@ def volume_content(tmpdir: Path) -> Path:
 @pytest.fixture
 async def mock_container_with_data(
     volume_content: Path, monkeypatch: pytest.MonkeyPatch
-) -> AsyncIterable[None]:
+) -> AsyncIterable[str]:
     async with aiodocker.Docker() as client:
         container = await client.containers.run(
             config={
@@ -56,7 +58,7 @@ async def mock_container_with_data(
         container_name = container_inspect["Name"][1:]
         monkeypatch.setenv("HOSTNAME", container_name)
 
-        yield None
+        yield container_inspect["Id"]
 
         await container.delete(force=True)
 
@@ -68,8 +70,24 @@ def downlaoded_from_s3(tmpdir: Path) -> Path:
     return path
 
 
+@pytest.fixture
+async def mock__get_self_container_ip(
+    mock_container_with_data: str,
+    mocker: MockerFixture,
+) -> None:
+    container_ip = await run_command_in_container(
+        mock_container_with_data, command="hostname -i"
+    )
+
+    mocker.patch(
+        "simcore_service_agent.services.backup._get_self_container_ip",
+        return_value=container_ip.strip(),
+    )
+
+
 async def test_backup_volume(
-    mock_container_with_data: None,
+    mock_container_with_data: str,
+    mock__get_self_container_ip: None,
     volume_content: Path,
     project_id: ProjectID,
     swarm_stack_name: str,