Delete channel or resources task fails when attempting to release an expired Redis lock

[radinamatic]

Observed behavior

Both Delete channel and Manage > Delete (all) resources for that same channel are failing.
Asset from v0.14.4-weekly7 build on Windows 7.
The only content present was KA (EN) > Math > Arithmetic, and I was not able to delete neither the full channel, nor by selecting all the resources inside, as the task was reported as Failed. I suspect just the DB was deleted but not the resources.

Expected behavior

Being able to delete channel/resources.

User-facing consequences

Inability to manage channels and resources on the device.

Errors and logs

Nothing jumped at me as relevant in the browser console, attaching the logs.

logs.zip

Steps to reproduce

1. Download KA (EN) > Math > Arithmetic.
2. Try to delete the channel and/or all the resources.

Context

Tell us about your environment, including:

• Kolibri version: v0.14.4-weekly7
• Operating system: Windows 7
• Browser: Firefox

[rtibbles]

Seems to be an issue here:

  File "c:\python36\lib\site-packages\kolibri\core\tasks\worker.py", line 72, in handle_finished_future
    result = future.result()
  File "c:\python36\lib\concurrent\futures\_base.py", line 425, in result
    return self.__get_result()
  File "c:\python36\lib\concurrent\futures\_base.py", line 384, in __get_result
    raise self._exception
  File "c:\python36\lib\concurrent\futures\thread.py", line 56, in run
    result = self.fn(*self.args, **self.kwargs)
  File "c:\python36\lib\site-packages\kolibri\core\tasks\worker.py", line 217, in wrap
    return f(*args, **kwargs)
  File "c:\python36\lib\site-packages\kolibri\core\tasks\job.py", line 194, in y
    result = func(*args, **kwargs)
  File "c:\python36\lib\site-packages\kolibri\dist\django\core\management\__init__.py", line 131, in call_command
    return command.execute(*args, **defaults)
  File "c:\python36\lib\site-packages\kolibri\dist\django\core\management\base.py", line 330, in execute
    output = self.handle(*args, **options)
  File "c:\python36\lib\site-packages\kolibri\core\tasks\management\commands\base.py", line 119, in handle
    return self.handle_async(*args, **options)
  File "c:\python36\lib\site-packages\kolibri\core\content\management\commands\deletecontent.py", line 114, in handle_async
    channel, node_ids, exclude_node_ids, force_delete
  File "c:\python36\lib\site-packages\kolibri\core\content\management\commands\deletecontent.py", line 45, in delete_metadata
    channel.delete_content_tree_and_files()
  File "c:\python36\lib\site-packages\kolibri\core\utils\cache.py", line 70, in __exit__
    self.release()
  File "c:\python36\lib\site-packages\kolibri\core\utils\cache.py", line 64, in release
    self._lock.release()
  File "c:\python36\lib\site-packages\kolibri\dist\diskcache\recipes.py", line 165, in release
    assert is_owned, 'cannot release un-acquired lock'
AssertionError: cannot release un-acquired lock

This is called after the channel metadata has been removed from the DB, but before the content is cleared up, so it's consistent with the reports.

[jredrejo]

FYI, in a RPI, with kolibri-server and this setup for redis:

CACHE_BACKEND = redis
CACHE_REDIS_MAXMEMORY = 536870912
CACHE_REDIS_MAXMEMORY_POLICY = allkeys-lru

deleting the same channel @radinamatic comments has worked, but... it took 8 minutes

[radinamatic]

Still extant on 0.14.4-RC1.

I only see the AssertionError in a warning, but there are ...to many SQL variables errors.

daemon.txt
kolibri.txt

[rtibbles]

Looks like my updates didn't go far enough here.

[radinamatic]

Additional detail: above fail happened when I selected to delete all the resources (Manage > select topic > Delete), but the workflow Options > Delete channel > select channel > Delete worked and wiped it in around 3 minutes.

[jonboiser]

Closing and opening #7697 with more detailed documentation