Contemporary Computing Technologies for Processing Big Spatiotemporal Data - Space-Time Integration in Geography and GIScience - page 327

Fig. 18.2 GEOSS index kernel

SQL Azure, which is a cloud-based database service provided by Microsoft.
The Postgresql version of the GEOSS clearinghouse is persistently maintained
and hosted on server and cloud computing platform. Postgresql database with its
PostGIS spatial extension provides high performance search capabilities for both
spatial information and text information, management of big data exceeding 4
terabytes, and support to concurrent user accesses.
A properly designed indexing mechanism can significantly improve big data
discovery by supporting high-speed data retrieval process (Theodoridis et al. 1996 ).
In general, GEOSS clearinghouse indexes text description of the features (such as
natural resources and national hydrology) and geographic location and temporal
extension of features and boundaries. Figure 18.2 shows the database and index
framework of GEOSS clearinghouse. Białecki et al. ( 2012 ) is used to index one
dimension text-based data. Spatial index mechanism is supported by Geotools or
Postgis spatial extension. By adopting the index mechanism, GEOSS clearinghouse
can support high-speed data retrieval process for data discovery.
Maintaining a consistently quick response time is critical, especially when
GEOSS clearinghouse needs to store millions to billions of metadata. The spatial
and temporal dimensions of the data cause this problem. Research shows that
if the response time is longer than 3 s, the users will become frustrated (Nah
2004 ). Besides adopting index mechanisms, GEOSS clearinghouse also employed
buffering and multi-threading mechanisms. Based on the above strategies, the
GEOSS clearinghouse provides consistently quick responses (1-2 s) for concurrent
searches from different domain users.