Refactor generic cleanup mechanism part 3

Current cleanup mechanism is awful. It's hardcoded, without repeat on failures,
contains a lot of mistakes, in some places it is too fast (in case of deletion of VMs)
in other to slow (like deletion of users). As well there is mix of cleanup mechanism
and resource cleanup mechanism.
To resolve these all issues this patch introcude a cleanup engine that resolves all issues
above.
It's resource based, so to add new resouce you should just a make subclass of base.Resource
and probably override some of methods like (list, delete, is_deleted) and that's all.
All complexity of managing:
0) waiting until async deletion is finished
1) repeat on failure logic
2) greaceful failure handling
3) parallelization
4) plugin support
Is hidden deep inside cleanup engine
+ bonus we are able to specify now single resource (without clenaping whole service)

PART 3:
-------

*) Refactor all contexts cleanups method to use new generic cleanup engine
   insted of cleanup.utils
*) Remove obsolate cleanup.utils
*) Fix all tests

bp benchmark-context-cleanup-refactor

Change-Id: I70557e6ebb56bbe565792d9ee854d3e78428a881

doc/feature_request/production_ready_cleanup.rst

@@ -11,15 +11,24 @@ Rally should delete in any case all resources that it created during benchmark.
 Problem Description
 -------------------
 
-* Deletion rate limit
+* (implemented) Deletion rate limit
 
   You can kill cloud by deleting too many objects simultaneously, so deletion
   rate limit is required
 
-* Retry on failures
+* (implemented) Retry on failures
 
   There should be few attempts to delete resource in case of failures
 
+* (implemented) Log resources that failed to be deleted 
+
+  We should log warnings about all non deleted resources. This information
+  should include UUID of resource, it's type and project.
+
+* (implemented) Pluggable
+
+  It should be simple to add new cleanups adding just plugins somewhere.
+
 * Disaster recovery
 
   Rally should use special name patterns, to be able to delete resources