Persistent Storage Modelling

gistfile1.txt

Some notes on Persistent Storage models.

Storage models do not require accurate details of physical storage technologies. 

Storage Capacity Model
This is a simple model representing the abstraction of persistent storage. The model must be easy to understand as a abstract model 
while scaling to support different types of “addressable storage” subsystems, ranging from cloud/enterprise storage technologies (SAN, 
NAS, DSaaS, etc) down to the smallest deployment unit of 'an IoT device with one internal disk'.

gistfile2.txt

Compute Workload Model
This is a simple model representing the abstraction of “IO workload streams”, used to generating "Addressable Storage Capacity" models.
Workload profiles with 'aggregate IO' as the scaling unit supports the broadest range of system profiles. 
This model is based on standard definitions of IO streams.

gistfile3.txt

Storage Management Model (aka OpenSDS)
This is a management model allowing storage systems to reliably address, classify, query and control a core subset of addressable
logical storage objects. 

gistfile4.txt

Data storage as a Service (DaaS)
This is a simple model representing the presentation of “addressable storage units” to virtualized execution environments.
It can be considered a de-facto definition for “Cloud Storage” where the NAS subsystem is the Service Provider.

gistfile5.txt

Persistance Model:
This is the list of "addressible persistent storage units" whose data must be persisted to an external backup storage.

Storage Capacity Model