chapter 13

Question 1

What is a distributed DBMS?

Answer 1

data is distributed across multiple systems
distribution by design
4 layer schema architecture: external schema, global conceptual schema, local conceptual schema, local internal schema
Distribution strategies: horizontal partitioning, vertical partitioning, partial replication (to improve performance).
tightly coupled. Heterogenity is not an issue.

Question 2

What is a federated DBMS?

Answer 2

Its a DBMS based on a global, federated conceptual schema. It describes integrated view of all participating data sources using the canonical data model. It realizes distribution transparency and application can access multiple data sources withing the same query.

Question 3

What are the alternatives for federated schema creation?

Answer 3

Bottom-up: schema integration

top-dowm: schema design, schema mapping

Question 4

What is the schema reference architecture of a federated schema?

Answer 4

External schema, federated schema, export schema, component schema, local schema

Question 5

Explain federated schema:

Answer 5

Includes (possibly just renames) export schema elements. May provide integrated schema (resolve structural, schematic and semantic heterogeneity).

Question 6

Explain export schema:

Answer 6

Describes subset of local data/schema to be made available for global application
May be under the control of the local system

Question 7

Explains: components schema:

Answer 7

describes the local DB usign global data model. Overcome data model heterogeneity

Question 8

Explains local schema:

Answer 8

Corresponds to the conceptual schema of the local DB.

It is based on local data model

Question 9

Mediator-based information systems

Answer 9

It’s a generalization of federated DBMS based on wrappers (for accessing data sources) and mediators (to access one or more wrappers and provide useful services)

Question 10

Components mediator-based IS:

Answer 10

external schema; global (mediator) schema;wrapper schema; local schema

Question 11

What are the tasks of the wrapper?

Answer 11

Encapsulate data source and provide uniform access to data. It provides infrastructure for overcoming heterogeneity among data sources: wrapper achitecture and wrapper interfaces
It helps to overcome hterogeneity of data sources regarding: data model, data acces api, query language and capabilities.
It gives support to global query evaluation and optimization.
Finally, it can be a useful infrastructure for Federated DBMS

Question 12

Garlic

Answer 12

„Wrapper Architecture“

Question 13

Wrapper architecture:

Answer 13

Garlic and wrapper cooperate for query processing (wrapper says its processing capabilities and garlic query engine compensates for lack of wrapper query functionality)

Question 14

Garlic - How do you model data as object collections?

Answer 14

The wrappers supplies description of data source in garlic data language when it registries itself in garlic.
Then a garlic object exists, with an interface, at least one implementation , identity and root objects (collections)

Question 15

Garlic - methods. How to call?

Answer 15

The methods can be called by garlic engine or by the application based on an object reference.

Question 16

Garlic - what are the mechanisms to invoke an method?

Answer 16

stub dispatch (natural if data source provides object class libraries)
Generic dispatch (wrapper provides one entry point and it is schema independent)

Question 17

Garlic - query planning. How it works?

Answer 17

Garlic optimizer builds the query plan using a bottom-up approach. the wrappers participate in query planning process.

Question 18

What are the methods the wrapper provides to be used by Garlic work requests?

Answer 18

Plan_access()
Plan_join()
Plan_bind()

Question 19

What is the result of a work request?

Answer 19

Sets of plans
Each plan contains a list of properties that describe which parts of the work request are implemented by the plan and what the costs are.

Question 20

What are the components the wrapper packeging provides in a package?

Answer 20

Interface facile (GDL definitions), enviroment files, libraries

Question 21

What is a foreign data server?

Answer 21

It is a server that manages data soreted outside the SQL server. In this case SQL server and SQL client use foreign server descriptors to communicatte with foreign servers.

Question 22

What is a foreign table?

Answer 22

it’s atable sotred in a foreign server or dynamically generated by foreign wrapper capabilities

Question 23

A multi-database system

loose coupling of systems.

Answer 23

It has no global schema (the actual integration needs to be performed by the application)
It has multi-database language.

Question 24

what are the limitations of SQL?

Answer 24

STANDARD SQL is unable to generically solve most forms of schematic heterogeneity.

Question 25

Define Schematic Query Languages

Answer 25

It is a solution that extends SQL to be able to transform data to metadata (and v.v.)
In additional, schematic query languages provide mechanism to access different databases in a single query

Question 26

What are the 6 operators of FIRA/FISQL?

Answer 26

????