How to connect to HDFS Server and fetch data in Statistica
Article ID: KB0082703
Updated On:
| Products | Versions |
|---|---|
| Spotfire Statistica | 13.0 and later versions |
Description
Article details how connections can be made to HDFS Server and how data can be fetched to Statistica and exported from Statistica to HDFS.
PRE-REQUISITES:
The URL to HDFS server and the credentials
Appropriate permissions in Enterpprise manager
PRE-REQUISITES:
The URL to HDFS server and the credentials
Appropriate permissions in Enterpprise manager
Issue/Introduction
This article details how to connect to HDFS Server and import and export data to HDFS into/from TIBCO Statistica
Environment
Windows
Resolution
1. Launch Enterprise Manager as a user who has effective Permissions : Database Admin (EXTDB_ADM)
2. Right Click on "Hadoop Distributed File Systems" and select New HDFS Server
3. Name the connection appropriately and enter username for the HDFS Server.
4. Click on Test connection to ensure the server is a HDFS Server
5. Define access permissions for the connection and click Commit to save the HDFS Server
6. Launch Statistica and open a new workspace
7. Use the HDFS Import Text or HDFS Export Text to either read from or write data to HDFS.
Here is a simple example :
1. In a new workspace, insert the HDFS Import Text node by either typing in Feature Finder or by adding it from Data | Manage(section) | External Data | HDFS Import Text
2. Click on the cog wheel in the top left hand corner of the node to open the parameters of the node.
3. Define options such as what delimiter is used for the file and other options such as if Variable names are on the first line, if blank lines should be skipped etc.
4. Click on File to Open HDFS Browser and select the HDFS Server from Servers listed , browse to the file you want to import from HDFS Server
5. Preview the file and click OK
5 . Run the node by clicking on the green button on the bottom left corner of the node.
6. Add other analytic nodes as needed (Optionally)
7. Insert HDFS Export node and open the parameters by clicking on cog wheel in the top left hand corner of the node
8. Select the server and click on File and choose the directory and filename to output the file
9. Run the node/ workspace by clicking Run All on the top corner of the workspace.
2. Right Click on "Hadoop Distributed File Systems" and select New HDFS Server
3. Name the connection appropriately and enter username for the HDFS Server.
4. Click on Test connection to ensure the server is a HDFS Server
5. Define access permissions for the connection and click Commit to save the HDFS Server
6. Launch Statistica and open a new workspace
7. Use the HDFS Import Text or HDFS Export Text to either read from or write data to HDFS.
Here is a simple example :
1. In a new workspace, insert the HDFS Import Text node by either typing in Feature Finder or by adding it from Data | Manage(section) | External Data | HDFS Import Text
2. Click on the cog wheel in the top left hand corner of the node to open the parameters of the node.
3. Define options such as what delimiter is used for the file and other options such as if Variable names are on the first line, if blank lines should be skipped etc.
4. Click on File to Open HDFS Browser and select the HDFS Server from Servers listed , browse to the file you want to import from HDFS Server
5. Preview the file and click OK
5 . Run the node by clicking on the green button on the bottom left corner of the node.
6. Add other analytic nodes as needed (Optionally)
7. Insert HDFS Export node and open the parameters by clicking on cog wheel in the top left hand corner of the node
8. Select the server and click on File and choose the directory and filename to output the file
9. Run the node/ workspace by clicking Run All on the top corner of the workspace.
Feedback
Yes
No
Powered by
