-
Type:
Bug
-
Status: Resolved
-
Priority:
Major
-
Resolution: Fixed
-
Affects Version/s: 2.1.1130
-
Fix Version/s: 2.4.6
-
Component/s: Local watcher
-
Tags:
-
Backlog priority:800
-
Sprint:nxMS 9.2.5, nxMS 9.2.6
-
Story Points:2
Root Cause:
When drive is offline due to network down, still the watcher threads are active.
localWatcher detects more than 200 events, which is higher than _windows_queue_threshold 50, this leads to enabling trigger_local_scan.
In local scan _scan(), we need the fix similar to NXDRIVE-644.
local watcher execute function()
while (1):
self._interact()
sleep(0.01)
if trigger_local_scan:
self._action = Action("Full local scan")
self._scan()
trigger_local_scan = False
self._end_action()
while (not self._watchdog_queue.empty()):
# Dont retest if already local scan
if not trigger_local_scan and self._watchdog_queue.qsize() > self._windows_queue_threshold:
log.debug('Windows queue threshold exceeded, will trigger local scan: %d events', self._watchdog_queue.qsize())
trigger_local_scan = True
self._delete_events.clear()
self._folder_scan_events.clear()
self._watchdog_queue = Queue()
break
evt = self._watchdog_queue.get()
self.handle_watchdog_event(evt)
self._win_delete_check()
self._win_folder_scan_check()
self._win_delete_check()
self._win_folder_scan_check()
Fix: introduce to_pause flag in _scan() similar to fix in NXDRIVE-644
def _scan(self):
log.debug("Full scan started")
start_ms = current_milli_time()
to_pause = not self._engine.get_queue_manager().is_paused()
if to_pause:
self._suspend_queue()
self._delete_files = dict()
self._protected_files = dict()
info = self.client.get_info(u'/')
self._scan_recursive(info)
self._scan_handle_deleted_files()
self._metrics['last_local_scan_time'] = current_milli_time() - start_ms
log.debug("Full scan finished in %dms", self._metrics['last_local_scan_time'])
self._local_scan_finished = True
if to_pause:
self._engine.get_queue_manager().resume()
self.localScanFinished.emit()
