This was implemented using the previously created Helpers.

This required to pass an AgentReference through to all garbageCollect() member functions of all objects.
The changed behavior required adaptation of the corresponding unit test (the archive queues now need to
be properly referenced in the root entry).

This field had a meaning when tapepools where stored in the objectstore and the archive queue a static object.
Since the introduction of the catalogue, they are now transient objects and storing their address
is not meaninful anymore.

Factored a first function out of OStoreDB into Herlpers.

The demangling has been ported in other places where RTTI was also used.

Adapted unit tests to validate this fetching.

Simplified handling of already (or soon to be) mounted tapes.
Added support for next mounts.
Added logging of decision steps.

Added many conditional printouts in unit tests. Those printouts are controlled by #defines in a central file.

Fixed calls to promise::get_future() after possible access form other thread. They are now guaranteed to happen before.
Added helgrind annotations for promise based synchronisation.
Added macros enabling helgrind annotations for shared_ptr.
Added suppression for shared_ptr used inside other standard lib object and not covered by the previous macros.
Added unit test for lower level .
Added suppressions for reported race conditions in Rados library.
Review heavily MemArchiveQueue and fixed missing commit in object store, leading to potentially orphaned objects.
Enabled formerly disabled test as it is now fast enough.

Lambdas were stored on stack and passed to async. They are now stored in object with proper lifetime.
Improved ArchiveMount::getNextJobBatch(): we now clean up the empty queues.

The ArchiveJobs are now updated using asynchronous IO.

...when it only contains orphaned references.

Added mechanism to make sure the ArchiveQueue lock is released only after all requests
have had their ownership updated.

AgentReference::queueAndExecuteAction() missed the waiting on the promise.
Added log when a non-owned job is popped from an archive queue.

AgentReference::{addTo|RemoveFrom}Ownership() are used instead.
Fixed typo in function name.
Added missing lock in AgentReference::queueAndExecuteAction().

Those queues will hold in memory the addition of the archive requests from several threads
and them all in one go, in order to limit contention and round trips on the object store.

Fix to #47: OStoreDB::getArchiveJobs() now returns an empty list for tape pools unknown to object store.

The scheduling based on disk throughput estimates will not be implemented in the forseeable, if ever.

Added drive status reporting in scheduler
Added drive status reporting in OStoreDB
Added support for drive status listing in the front end
Removed virtual functions from the Scheduler, which is never overloaded.
Added DesiredDriveState structure to drive state.
Removed usage of duplicate MountType, DriveStatus and DriveState structures.
Created DriveInfo structure to allow recreation of drive register entry in all
reporting situation (potentially with partial/assumed info).

Agent becomes a usual class, like any other in the object store.
Object name generation (agent's and others) is now handled in the separate AgentReference.
The previous situation created race conditions.

Completed changes so that archive and retrieves can be queued, mounts created and jobs reported done.

Fixed archive queue not being deleted on last archive job pop.
Next step is retrieve mount scheduling.