4. Transparency Goals

Question 1

What does transparency mean?

Answer 1

A process or interface functioning without the user being aware of its presence

Question 2

What are the standard set of transparencies?

Answer 2

Access: hiding how a resource is accessed
Location: hiding where a resource is located
Migration: hiding the fact that a resource may move location
Relocation: hiding the fact a resource may move to another location during use
Replication: Hiding the fact a resource is replicated
Concurrency: Hiding the fact a resource may be shared by several users
Failure: hiding the failure and recovery of a resource

Question 3

Why is transparency desirable?

Answer 3

Allows us to build DSs “as if” they were centralised. Or “as if” the axioms weren’t in force

Question 4

How do we achieve transparency of location?

Answer 4

Transport and internet protocols (DNS). Naming, addressing, connection and routing

Question 5

How do we achieve transparency of migration?

Answer 5

Dynamically update directories of hosts and routes.

Question 6

How do we achieve transparency of relocation?

Answer 6

The physical and link layers take care of handover from cell to cell. Transport and internet protocols are still in charge

Question 7

How do we achieve transparency of replication?

Answer 7

Transport and internet protocols are not enough here, distributed processors and resource managers need to schedule cleverly

Question 8

How do we achieve transparency of access?

Answer 8

Use interoperable standards XML, RDF etc. Compensate with wrappers and mappers to build mediator middleware.

Question 9

How do we achieve transparency of concurrency?

Answer 9

Resource/ process managers handle concurrent access using the following algorithms:
selecting a coordinator
clock synchronisation
joint decision making
May lead to issues such as deadlock and livelock.

Question 10

How do we achieve transparency of failure?

Answer 10

Sender can poll/ try again