Decrease time_threshold for zone_purge task.

[Environment]

Ussuri
Charms 20.08

[Description]

After deleting a zone on a designate-bind backend, the zone
remains active until the zone purge producer task gets executed.

$ openstack zone show 98c02cfb-5a2f-xxxx
Could not find Zone

$ openstack zone delete 98c02cfb-5a2f-xxxx
Could not find Zone

mysql> select * from zones where id="98c02cfb-5a2f-xxxx";
Empty set (0.01 sec)

363:2020-09-25 05:23:41.154 1685647 DEBUG designate.central.service [req-8223a934-84df-44eb-97bd-a0194343955a - - - - -] Performing purge with limit of 100 and criterion of {u'deleted': u'!0', u'deleted_at': u'<=2020-09-18 05:23:41.143790', u'shard': u'BETWEEN 1365,2729'} purge_zones /usr/lib/python2.7/dist-packages/designate/central/service.py:1131

No zone was found by this criteria, therefore, the hard delete zone from the database https://github.com/openstack/designate/blob/2e3d8ab80daac00bad7d2b46246660592163bf17/designate/storage/impl_sqlalchemy/__init__.py#L454
didn't apply.

This delta is governed by time_threshold https://github.com/openstack/designate/blob/89435416a1dcb6df2a347f43680cfe57d1eb0a82/designate/conf/producer.py#L100
which is set to 1 week.

### Proposed actions

1) Make the time_threshold shorter to a 1 hour span by default.
2) Don't list deleted_at zones in designate list operation.