ceph/qa/workunits/cephadm/test_cephadm_timeout.py

180 lines
5.6 KiB
Python
Executable File

#!/usr/bin/python3 -s
import time
import os
import fcntl
import subprocess
import uuid
import sys
from typing import Optional, Any
LOCK_DIR = '/run/cephadm'
DATA_DIR = '/var/lib/ceph'
class _Acquire_ReturnProxy(object):
def __init__(self, lock: 'FileLock') -> None:
self.lock = lock
return None
def __enter__(self) -> 'FileLock':
return self.lock
def __exit__(self, exc_type: Any, exc_value: Any, traceback: Any) -> None:
self.lock.release()
return None
class FileLock(object):
def __init__(self, name: str, timeout: int = -1) -> None:
if not os.path.exists(LOCK_DIR):
os.mkdir(LOCK_DIR, 0o700)
self._lock_file = os.path.join(LOCK_DIR, name + '.lock')
self._lock_file_fd: Optional[int] = None
self.timeout = timeout
self._lock_counter = 0
return None
@property
def is_locked(self) -> bool:
return self._lock_file_fd is not None
def acquire(self, timeout: Optional[int] = None, poll_intervall: float = 0.05) -> _Acquire_ReturnProxy:
# Use the default timeout, if no timeout is provided.
if timeout is None:
timeout = self.timeout
# Increment the number right at the beginning.
# We can still undo it, if something fails.
self._lock_counter += 1
start_time = time.time()
try:
while True:
if not self.is_locked:
self._acquire()
if self.is_locked:
break
elif timeout >= 0 and time.time() - start_time > timeout:
raise Exception(self._lock_file)
else:
time.sleep(poll_intervall)
except Exception:
# Something did go wrong, so decrement the counter.
self._lock_counter = max(0, self._lock_counter - 1)
raise
return _Acquire_ReturnProxy(lock=self)
def release(self, force: bool = False) -> None:
if self.is_locked:
self._lock_counter -= 1
if self._lock_counter == 0 or force:
self._release()
self._lock_counter = 0
return None
def __enter__(self) -> 'FileLock':
self.acquire()
return self
def __exit__(self, exc_type: Any, exc_value: Any, traceback: Any) -> None:
self.release()
return None
def __del__(self) -> None:
self.release(force=True)
return None
def _acquire(self) -> None:
open_mode = os.O_RDWR | os.O_CREAT | os.O_TRUNC
fd = os.open(self._lock_file, open_mode)
try:
fcntl.flock(fd, fcntl.LOCK_EX | fcntl.LOCK_NB)
except (IOError, OSError):
os.close(fd)
else:
self._lock_file_fd = fd
return None
def _release(self) -> None:
fd = self._lock_file_fd
self._lock_file_fd = None
fcntl.flock(fd, fcntl.LOCK_UN) # type: ignore
os.close(fd) # type: ignore
return None
def _is_fsid(s):
try:
uuid.UUID(s)
except ValueError:
return False
return True
def find_fsid():
if not os.path.exists(DATA_DIR):
raise Exception(f'{DATA_DIR} does not exist. Aborting...')
for d in os.listdir(DATA_DIR):
# assume the first thing we find that is an fsid
# is what we want. Not expecting multiple clusters
# to have been installed here.
if _is_fsid(d):
return d
raise Exception(f'No fsid dir found in {DATA_DIR} does not exist. Aborting...')
def main():
print('Looking for cluster fsid...')
fsid = find_fsid()
print(f'Found fsid {fsid}')
print('Setting cephadm command timeout to 120...')
subprocess.run(['cephadm', 'shell', '--', 'ceph', 'config', 'set',
'mgr', 'mgr/cephadm/default_cephadm_command_timeout', '120'],
check=True)
print('Taking hold of cephadm lock for 300 seconds...')
lock = FileLock(fsid, 300)
lock.acquire()
print('Triggering cephadm device refresh...')
subprocess.run(['cephadm', 'shell', '--', 'ceph', 'orch', 'device', 'ls', '--refresh'],
check=True)
print('Sleeping 150 seconds to allow for timeout to occur...')
time.sleep(150)
print('Checking ceph health detail...')
# directing stdout to res.stdout via "capture_stdout" option
# (and same for stderr) seems to have been added in python 3.7.
# Using files so this works with 3.6 as well
with open('/tmp/ceph-health-detail-stdout', 'w') as f_stdout:
with open('/tmp/ceph-health-detail-stderr', 'w') as f_stderr:
subprocess.run(['cephadm', 'shell', '--', 'ceph', 'health', 'detail'],
check=True, stdout=f_stdout, stderr=f_stderr)
res_stdout = open('/tmp/ceph-health-detail-stdout', 'r').read()
res_stderr = open('/tmp/ceph-health-detail-stderr', 'r').read()
print(f'"cephadm shell -- ceph health detail" stdout:\n{res_stdout}')
print(f'"cephadm shell -- ceph health detail" stderr:\n{res_stderr}')
print('Checking for correct health warning in health detail...')
if 'CEPHADM_REFRESH_FAILED' not in res_stdout:
raise Exception('No health warning caused by timeout was raised')
if 'Command "cephadm ceph-volume -- inventory" timed out' not in res_stdout:
raise Exception('Health warnings did not contain message about time out')
print('Health warnings found succesfully. Exiting.')
return 0
if __name__ == '__main__':
if os.getuid() != 0:
print('Trying to run myself with sudo...')
os.execvp('sudo', [sys.executable] + list(sys.argv))
main()