|
|
It's a CONTROL issue!
|
|
Disparate data and information
integration, sharing and interoperability is all about CONTROL
over access, indexing, query processing, and result-set data:
 |
Data warehouses
gain control by moving data to a "one-size-fits-all" centralized
repository. The data source owner loses control over their
data. |
|
Conventional
adapters in federated systems have been described as
"virtual data
warehouses", where processes similar to those involved with data warehouses are accommodated
to a much lesser extent using adapters that attempt to gain some control over the query
process. However, query performance and index types are mainly
out of the control of conventional adapters in federated systems and under
control of data source systems. Conventional adapters are usually
custom-designed for every data source, queries tend to be
"hard-wired", and can involve a data schema transform to the underlying
data source schema. Access controls can be put in place. |
|
EIQ Products
assume control by:
|
Cleaning up data as it is read to build and
maintain indexes |
|
Using multiple indexes and
types, and applying consistent indexes across multiple
disparate data sources |
|
Enabling highly flexible
data models to be used for query generation and result-set
data
integration |
|
Indexing data and
processing all queries 100% in the EIQ Product indexes - no interim tables and no
interaction with, or load on, data sources |
|
Data source owners not
aware of indexing or query processing in middleware |
|
Data remains stored in data
source and only final result-set data is retrieved using
record-level pointers |
|
Indexes are based on underlying source data schemas OR on result-sets from data source systems |
|
Using the latest data
available in the data source |
|
Imposing a rigorous
role-based access control system from user-level to data
source row, column or data element-level |
|
Data warehouses and
conventional federated
adapters each have their role in an organization, but
when data cannot be stored elsewhere, systems are unable to support
additional indexes and queries, implementation time (and/or cost) is critical, and/or the
disparate nature of the data precludes a "one-size-fits-all"
approach, EIQ Products fulfill a unique role and provides control
in seemingly impossible situations.
|
|
|
