[itdssbuild01] CTA_build > cta-xrootd_plugins-fakeeos localhost 10955 '/my_dir?a=apple&b=banana'
Hello World
status.ToStr()=[SUCCESS]
status.IsError()=false
status.IsFatal()=false
status.IsOK()=true
[itdssbuild01] CTA_build > tail -n 1 /var/log/cta/cta-frontend.log
Feb 23 19:37:35 itdssbuild01 cta-frontend: <158>2017-02-23T19:37:35.584734+01:00 cta-frontend[23717]: LVL="Info" TID="23730" MSG="FSctl called" arg1="/my_dir" arg2="a=apple&b=banana" client->host="localhost" client->name="smurray"
[itdssbuild01] CTA_build >

This timeout actually covers a grace period for draining the metadata to central storage.
Bumped it up from 5 seconds to 5 minutes.

Refactoring layers to limit their number and catch possible errors using && instead of ; between commands

The previous case, keeping a reference created a race condition as 2 threads would
use the same logContext.

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.

Increased queue depth in helgrind reports.
Added locking in a constructor to make helgrind happy.
Added suppression for helgrind finding races in synchronization functions (?)

The timeout is now set to zero in unit tests.

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

The queue is housed in the AgentReference class. It currently covers the add to ownership
and remove from ownership operations. This queued system guarantees in-order execution
of operations.