Make namespace deletion asynchronous

Instead of deleting a namespace synchronously as before, we now mark the namespace for deletion, disable it, and rename it. A worker then comes along and deletes the namespace in the background. This results in a *significantly* better user experience, as the namespace deletion operation now "completes" in under a second, where before it could take 10s of minutes at the worse.

Fixes https://jira.coreos.com/browse/QUAY-838

initdb.py

@@ -20,7 +20,8 @@ from data.database import (db, all_models, beta_classes, Role, TeamRole, Visibil
                            ExternalNotificationEvent, ExternalNotificationMethod, NotificationKind,
                            QuayRegion, QuayService, UserRegion, OAuthAuthorizationCode,
                            ServiceKeyApprovalType, MediaType, LabelSourceType, UserPromptKind,
-                           RepositoryKind, TagKind, BlobPlacementLocation, User)
+                           RepositoryKind, TagKind, BlobPlacementLocation, User,
+                           DeletedNamespace)
 from data import model
 from data.queue import WorkQueue
 from app import app, storage as store, tf
@@ -893,8 +894,11 @@ def populate_database(minimal=False, with_storage=False):
     model.repositoryactioncount.update_repository_score(to_count)
 
 
+WHITELISTED_EMPTY_MODELS = ['DeletedNamespace']
+
 def find_models_missing_data():
-  # As a sanity check we are going to make sure that all db tables have some data
+  # As a sanity check we are going to make sure that all db tables have some data, unless explicitly
+  # whitelisted.
   models_missing_data = set()
   for one_model in all_models:
     if one_model in beta_classes:
@@ -903,7 +907,8 @@ def find_models_missing_data():
     try:
       one_model.select().get()
     except one_model.DoesNotExist:
-      models_missing_data.add(one_model.__name__)
+      if one_model.__name__ not in WHITELISTED_EMPTY_MODELS:
+        models_missing_data.add(one_model.__name__)
 
   return models_missing_data