mirror of
https://github.com/jhd3197/ServerKit.git
synced 2026-10-30 00:27:11 +09:00
Juan Denis
ec8cd9462e
ps, inspect, info, version, tag and the network/volume inspects ran with timeout=None, so a wedged dockerd held every request thread that asked, forever; the panel runs a single worker process. They now give up after 30s, and the Containers page reports Docker unreachable. Pull, build, logs and compose stay unbounded on purpose.
31 lines
980 B
Python
31 lines
980 B
Python
"""Docker metadata reads are bounded.
|
|
|
|
ps / inspect / info / version ran with timeout=None, so a wedged dockerd
|
|
held every request thread that asked forever. Long operations (pull,
|
|
build, compose up, logs) stay unbounded on purpose.
|
|
"""
|
|
from unittest.mock import patch
|
|
|
|
import pytest
|
|
|
|
from app.services.docker_service import QUICK_TIMEOUT, DockerService
|
|
|
|
OK = {'success': True, 'output': '', 'stderr': '', 'error': None, 'returncode': 0}
|
|
|
|
|
|
@pytest.mark.parametrize('call', [
|
|
lambda: DockerService.list_containers(),
|
|
lambda: DockerService.get_container('abc'),
|
|
lambda: DockerService.get_docker_info(),
|
|
lambda: DockerService.is_docker_installed(),
|
|
lambda: DockerService.tag_image('a:1', 'a:2'),
|
|
])
|
|
def test_metadata_reads_pass_a_timeout(call):
|
|
with patch.object(DockerService, 'run', return_value=OK) as run:
|
|
call()
|
|
assert run.call_args.kwargs['timeout'] == QUICK_TIMEOUT
|
|
|
|
|
|
def test_quick_timeout_is_bounded():
|
|
assert 0 < QUICK_TIMEOUT <= 60
|