|
|
|
EIQ Server® is highly flexible and can be deployed in a number
of ways
EIQ Server can work with almost ALL data and information in
multiple systems in multiple
locations on multiple platforms, accommodate non-EIQ Server result
sets, AND scale through EIQ Server to EIQ Server communication |
Distributed, Parallel Processing
EIQ Server underlying
technology has always been able to work with multiple attached databases
simultaneously
and this concept has been extended to EIQ Server. When EIQ Server
queries are executed, EIQ Server indexes, NOT the actual data
sources, are queried simultaneously. The total number or records
and the number of records found in each data source can be
determined before result-sets are retrieved from the data sources;
this information can also be subject to retrieval rules and/or
allow a degree of interaction for query refinement.
Result-sets from all data
sources are combined, sorted and ranked according to
rules in the EIQ Server initiating the query. Alternatively,
for larger
result-sets and systems, ranking rules can be sent with the query,
and then applied at the local EIQ Server retrieving data from the
data source, to restrict the
amount of data retrieved and high grade result-sets.
Single-point, Universal and Uniform Index, Query, Retrieval
and Integration
WhamTech sees database and search technologies
converging, whereby databases will support more search functions
and search engines will support more database functions (to a
lesser extent). It is estimated that 85% of corporate data and
information resides outside of structured databases. Even the 15%
of data that is estimated to reside within structured databases
can be unstructured in format and reside in separate databases with dissimilar data schemas.
To cover the wide range of data sources, different
indexing approaches and result-set data and information retrieval
options are required:
 |
For structured data, unique ROWID or primary key
access can be used to directly retrieve data
records using SQL; this is the fastest way to retrieve data from
a database. |
|
For unstructured text search, a URL (uniform
resource locator), UNC (universal naming convention), or other
file reference can be used, e.g., on a document server. |
|
For semi-structured data, links similar to
unstructured data in combination with other referential data can
be used, e.g., row and column for a spreadsheet, row and tag, or
file position for
an XML document. |
|
For unstructured text search on structured data,
links similar to unstructured data with additional row-level
information can be used, e.g., database x, table y, ROWID z. |
EIQ Server is “one-stop” middleware that appears as
a database driver to applications and can be used to access ALL structured
database, unstructured and semi-structured data and information.
EIQ Server Deployment Architecture
EIQ Servers can call on
other EIQ Servers, and different EIQ Server configuration files
can be used for different applications, security access levels,
etc. Also, of note, EIQ Servers do not need to conform to a fixed
hierarchical structure; lower-level data sources can be direct
connected to higher-level EIQ Servers, bypassing intervening
levels. EIQ Server deployment is illustrated in Figure 2, below. A
more detailed diagram is available in Appendix 3.
Figure 2:
EIQ Server deployment with multiple structured databases/files, lower-level EIQ
Servers, unstructured data sources/text, and as an adapter |
|
|
|
|
