dataset registration process sergey sukhonosov, dr. sergey belov national oceanographic data centre,...
TRANSCRIPT
Dataset registration process
Sergey Sukhonosov, Dr. Sergey BelovNational Oceanographic Data Centre, Russia
Training course on establishment of the ODP regional node and data network for SNDM-Argentina,7 - 11 October 2013, Buenos Aires, Argentina
2
Registration stepsDesignRegistrationMaintenance
3
DesignData source storage type
DBMS – all types of data;Structured data files – forecasts, climate, small volumes of
real-time data;Objective files – images, documents, shape-files, …;Web-application – WMS;
Data source structure (data files, tables/views)
4
Data source storage type - DBMSSupported databases: MySQL, PostgreSQL, Oracle, MS
SQL Server (native JDBC driver)Table structure requirement: one parameter – one
column.Possibility to get the data from one or several tablesFilter the dataBenefits:Metadata is updating automaticallyCaching modeFast processing of large amounts of data.Flexible adjustment of the data.
5
Data source storage type – Structured data files
Supported formats: text files with separator or text files with fix column positions.
Structure must be ‘plain’, non-hierarchical
Options to link data files:Specify URL to HTTP/FTP server,Upload data file to Data Provider server using web
interface,Copy data file to local Data Provider file system
6
Data source storage type – Structured data files
Benefits:• Metadata is updating automatically• Caching mode• In case of using HTTP/FTP data source it’s enough to
update data files, Data Provider will synchronize data, update data cache (instances) and update dynamic metadata (temporal extent, geographic bounding box,…)
• Connect set of data files to one metadata (specifying all file names or using filename mask (e.g. “*.txt”))
7
Objective data filesSupported formats: any data format – Data Provider
don’t analyze the content and provide data “as is”. In case if several object data files should be linked to one
metadata, they can be zipped to archive. E.g. GIS data (shp, shx, dbf).
The ways to link objective data files:• specify URL to HTTP/FTP server,• upload data files to Data Provider server using web
interface
8
Objective data filesAt least one object data file must be linked to discovery
metadata. Linked data files call ‘instances’ and have their own discovery metadata.
Metadata for instance based on resource metadata.Features:• Once data file is uploaded it can’t be updated
automatically: only manually.• You should add new data file manually via web-interface• Dynamic metadata is not updating automatically by
scheduler
9
Design
Data granularity issue: a) Split data to several metadata records by geographic
box, time period, measured parameter;
10
Designb) Split data to data subsets (instances) by “key”
parameter;
11
Data granularity – point data
12
Data granularity – point data
2001-08-06T00:00:00;-7.778;119.603;V2AP6.00;0;27.3
2001-08-06T00:00:00;-7.778;119.603;V2AP6.00;68;27.2
2001-08-06T00:00:00;-7.778;119.603;V2AP6.00;92;26.4
2001-08-06T00:00:00;-7.778;119.603;V2AP6.00;100;25.6
2001-08-06T00:00:00;-30.515;-155.227;DCFH2.00;2;17.9
2001-08-06T00:00:00;-30.515;-155.227;DCFH2.00;4;17.9
2001-08-06T00:00:00;-30.515;-155.227;DCFH2.00;6;17.9
2001-08-06T00:00:00;-30.515;-155.227;DCFH2.00;8;17.9
13
Data granularity – profile data
14
Data granularity – profile dataSpecifying record header
15
Data granularity – profile dataSpecifying list of parameters in the data table
16
Data granularity – profile data
2001-08-06T00:00:00;-7.778;119.603;V2AP6.000;27.368;27.292;26.4100;25.6
2001-08-06T00:00:00; -30.515;-155.227;DCFH2.002;17.94;17.96;17.98;17.9
17
System parameters analysisMake cross-mapping between local data columns and
system parametersContact global node (MINCYT) to add new parameters if
required
18
RegistrationUse web-interface to register metadata and connect the
data (Notice: Chrome browser is not supported yet)Use metadata templatesSpecify maximum details to fit metadata quality
requirements
19
MaintenanceUpdate data in-timeUpdate metadata on data frequency basis – create
schedulers for automatic metadata actualizationAnalyze daily reports from Integration Server
20
Questions?