Difference between revisions of "Statistical Algorithms Importer: Publish Algorithms"
(→Publishing an algorithm) |
(→Publishing an algorithm) |
||
| Line 4: | Line 4: | ||
Publishing an algorithm means making it available as-a-Service on the e-Infrastructure, for some Virtual Research Environment (VRE). In turn, this requires the algorithm to undergo a coherence check with respect to the rest of the infrastructure and be shipped on the DataMiner machines. Only the JAR file will be deployed on the machines, leaving the scripts on the private Workspace of the user. This mechanism saves policy requirements of those users who do not want to share their code. Only the compiled Java code will be deployed on the DataMiner machines (inside the JAR file), thus the e-Infrastructure staff will not be able to get the scripts code. The compiled Java file contains a link to a Workspace location, accessible only via authentication (owned only by the DataMiner service). The service will download the scripts and resources on-the-fly, execute the main script and later delete everything. | Publishing an algorithm means making it available as-a-Service on the e-Infrastructure, for some Virtual Research Environment (VRE). In turn, this requires the algorithm to undergo a coherence check with respect to the rest of the infrastructure and be shipped on the DataMiner machines. Only the JAR file will be deployed on the machines, leaving the scripts on the private Workspace of the user. This mechanism saves policy requirements of those users who do not want to share their code. Only the compiled Java code will be deployed on the DataMiner machines (inside the JAR file), thus the e-Infrastructure staff will not be able to get the scripts code. The compiled Java file contains a link to a Workspace location, accessible only via authentication (owned only by the DataMiner service). The service will download the scripts and resources on-the-fly, execute the main script and later delete everything. | ||
| − | The software publication can be run by pressing the Publish button. At this point, a link to the JAR file and to the algorithm metadata will be sent via e-mail to the e-Infrastructure managers. Once the algorithm has been deployed, the manager will alert the user via the e-Infrastructure messaging system. | + | The software publication can be run by pressing the Publish button, be careful, you must always do Create operation before Publish. At this point, a link to the JAR file and to the algorithm metadata will be sent via e-mail to the e-Infrastructure managers. Once the algorithm has been deployed, the manager will alert the user via the e-Infrastructure messaging system. |
[[Image:StatisticalAlgorithmsPublish.png|thumb|center|800px|Statistical Algorithms Importer (SAI), portlet. Publish button.]] | [[Image:StatisticalAlgorithmsPublish.png|thumb|center|800px|Statistical Algorithms Importer (SAI), portlet. Publish button.]] | ||
[[Category:Statistical Algorithms Importer]] | [[Category:Statistical Algorithms Importer]] | ||
Revision as of 11:50, 19 July 2017
This page explains how to publish algorithms created using the Statistical Algorithms Importer (SAI).
Publishing an algorithm
Publishing an algorithm means making it available as-a-Service on the e-Infrastructure, for some Virtual Research Environment (VRE). In turn, this requires the algorithm to undergo a coherence check with respect to the rest of the infrastructure and be shipped on the DataMiner machines. Only the JAR file will be deployed on the machines, leaving the scripts on the private Workspace of the user. This mechanism saves policy requirements of those users who do not want to share their code. Only the compiled Java code will be deployed on the DataMiner machines (inside the JAR file), thus the e-Infrastructure staff will not be able to get the scripts code. The compiled Java file contains a link to a Workspace location, accessible only via authentication (owned only by the DataMiner service). The service will download the scripts and resources on-the-fly, execute the main script and later delete everything.
The software publication can be run by pressing the Publish button, be careful, you must always do Create operation before Publish. At this point, a link to the JAR file and to the algorithm metadata will be sent via e-mail to the e-Infrastructure managers. Once the algorithm has been deployed, the manager will alert the user via the e-Infrastructure messaging system.
