aql.cleanup() will hang if there are failed tasks.

[guohui-gao]

Describe the bug
I added aql.cleanup() to a dag to clear temp tables.
it will run after all task success. however, if there are failures in the tasks. it will hang, with the
following errors:

[2022-07-12, 09:19:35 UTC] {cleanup.py:103} INFO - waiting on the following tasks to complete before cleaning up: [('complete', 'upstream_failed'), ('tear_down_tasks', 'removed'), ('create_staging_comments', 'upstream_failed'), ('merge_comments', 'upstream_failed'), ('upload_comments_to_gcs', 'upstream_failed'), ('extract_finish', 'upstream_failed'), ('create_comments', 'upstream_failed'), ('load_comments_to_snowflake_staging_db', 'upstream_failed'), ('cleanup_staging_comments', 'upstream_failed')]
[2022-07-12, 09:19:35 UTC] {cleanup.py:137} WARNING - You are currently running the 'waiting mode', where the task will wait for all other tasks to complete. Please note that for asynchronous executors (e.g. SequentialExecutor and DebugExecutor, this mode will cause locks.

Version

• Astro: [e.g. 0.11.0]
• OS: [eg. Debian]

Expected behavior
Cleanup would run whether all tasks success or if there are any task fail.