Allow to set an alba fragment cache size limitation per vPool

[wimpers]

Currently an alba fragment cache is shared between all vPool that are configured to use that SSD backend. As each vPool is a different customer it would be good to limit the amount of space (in %) the vPool can consume in the fragment cache. This size limitation should be dynamically updateable so you don't have to restart the proxy in case new SSD are added to the backend.

Adding such a limitation would require:

• have a different prefix (for cache namespaces) per vpool
• periodically refresh how much space all cache namespaces with this prefix occupy
• do eviction per vpool

[wimpers]

Continued in https://github.com/openvstorage/alba_ee/issues/6