Páginas filhas
  • BI00002_Smart_Analytics_update_flow

Smart Analytics update flow

Produto:

TOTVS Smart Analytics

Versions:

11.80 and 12

Event:

 

How to update the project TOTVS Smart Analytics

 

Step by Step:

Introduction

Once in every two months releases with update packages of TOTVS Smart Analytics will be published, which will be available on the Customer Portal, and the client will be free to update the product according to its needs. Packages include the latest default project, with graphic modeling, graphs and Scripts MAQL of modeling, apart from Agent and its configuration.

This article mainly intends to instruct the steps that must be followed to update the whole project. To make the process easier, the route is divided in 2 possible scenarios:

  • Scenario 1: There is no customization, that is, the customer keeps TOTVS Smart Analytics default project.
  • Scenario 2: The customer customized the project, in order to meet its needs. In this case, the project maintenance and update process is of total responsibility of the user. But in this article, some points will be clarified to make this procedure easier.

The first step, for both scenarios, is to download the package with the new version of TOTVS Smart Analytics and the Agent package, save and extract the packages on the same folder where the older projects are kept.

Then, recommendations of good practices are listed, which must be executed before starting the following steps:

  • From the start moment of the update process, it must be completed until the end, involving modeling, ETL and Agent. If any step is incomplete, errors may occur.
  • It is crucial to remove Schedules (appointments) that are on GoodData, in order to not start a new load execution, before the update is completed, which might generate errors.
  • For Scenario 2: The user must remain using its own customized project, and edit it as the updates are added, instead of modifying the default project. It is also recommended to backup the customized project before starting.

Once these actions are completed, the user may continue the update process of TOTVS Smart Analytics. The process must always follow the same sequence: first, the update and release of modeling, followed by the project deploy, and finally, update of Agent configuration files.

 

 

     

    Modeling Backup and Update

    On scenario 1, it is strongly recommended to the user to backup the modeling, in order to avoid future errors that may occur, then publish the latest modeling.

    Modeling backup

    Follow the steps below:

      1. Access the model folder within the project structure, right-click in New / Logical Data Model.
      2. Name the file as you wish and click Finish.
      3. On the menu window Details (at right) click Import / GoodData Project.
      4. Select your project name GoodData, click OK and save the modeling.

    Modeling update

    After saving the backup, follow the steps below to update the modeling.

      1. At first, open the default project updated on Cloud Connect, following the steps below:
          1. At left, click on View Navigator / Import.
          2. On the dialog window, select General / Existing Projects into Workspace.
          3. Keep the option Select root directory, and click Browse.
          4. Select the path and click the project you wish. The project folder to be selected must have the following structure:
          5. Click OK.
          6. Back on the dialog box, click Finish.
          7. The project will be opened on the tool.
        1. Double-click the folder Model.
        2. Double-click the file: TOTVSSA.mdl.
        3. Click Publish model to server. In this moment, a comparison occurs between the content of GoodData cloud and what is on the local disk.
        4. After viewing the modifications (you may also view the script MAQL), click OK, to proceed.
        5. Wait for the message informing that the modifications were successful and click OK.

    ETL process update

    There are two ways to re-deploy the project and update the ETL process of TOTVS Smart Analytics. In one way, you must only access the project in analytics.totvs.com.br, on the other, it is crucial to use the Cloud Connect tool. These very same steps will be followed by users that are eligible for Scenario 2, after other procedures explained on the proper section.

    First method:

      1. Access the project you wish to re-deploy in analytics.totvs.com.br.
      2. No Menu superior clique em Manage e selecione Data Integration Console.     
      3. In Overview you may view all projects published.
      4. Select the published project you wish, click Re-deploy.  
      5. On the opened dialog window, click Browse, and select the compacted project where the download was performed on the Client Portal.

        Note

        If you are using this steps on Scenario 2, do not forget to compact the project again, with the customizations .

      6. In Process Type, keep Cloud Connect and in Process Name, keep the same name, preferably.  
      7. Click Re-Deploy and wait for the load to complete.

     

    Second Method (using Cloud Connect):

      1. To open the Cloud Connect project, follow the steps from "i" to "vii" illustrated on step A of item Modeling Update above.
      2. Access Server Explorer / GoodData Projects.
      3. Select the project you wish to re-deploy, select Set as Working Project for "Totvs Smart Analytics".    
      4. In this project selected, expand it clicking on the arrow at the left of the name. Expand the folder Processes again.
      5. The project that is already on the cloud will appear. Right-click it and select Re-deploy.  
      6. Wait for the message of successful re-deploy.

     

    Update of Agent configuration files

    my.properties

     

    To update the my.properties file, follow the steps below:

      1. Open the file my.properties, located on the package Agent, with a text program.
      2. Complete the information again regarding login, password, project ID, metadata ID and connection with Database (use the previous version, already completed, as reference).
      3. Upon completing, save the information.
    1. run.bat

    In both scenarios, check the status of your run.bat within Agent package. This file must be reference for the new version of the Agent within this very same package, and for my.properties that the user just configured with the steps above. In such cases, the previous version of the customer has extra parameters to improve performance, for example. It that is the case, add the parameter for run.bat updated.

    Following, two examples of the structure of the run.bat file. One example consists on general structure, the other, illustrates a parameter added in order to improve performance:

                 

     

     

    On scenario 2, the user is responsible for maintenance and update of its project. The update of a customized project demands extreme caution by the user, to not lose its customizations nor updates published, therefore there are some recommendations to follow:

    1. Do not publish modeling and the project re-deploy from the package downloaded from Portal, because the project published will be replaced and customizations will be lost.
    2. The first step must be Release Notes analysis (available on the Customer Portal) because all necessary information regarding the update released are there. From it on, the user may evaluate if what was update will interfere in its customizations.
    3. As previously warned, by proceeding with the update of the customized project, perform it on your local project, never overwrite the standard project of TOTVS Smart Analytics.

    Modeling

    To simplify the modeling update of users that have customizations, the TOTVS Smart Analytics package integrates MAQL scripts of modeling. Follow the next steps:

      1. Access https://analytics.totvs.com.br/gdc/md/<ID_PROJETO>/ldm/manage2

        Note

        Complete the field <ID_PROJETO> with the ID regarding its project.

      2. Open the MAQL scripts of modeling, copy the script and enter it on the proper space.
      3. Click Submit.
      4. To continue, click the access link.
      5. Wait for the successful message. 

     

    Warning

    If the user does not keep updates as the latest, the MAQL Script of the last publish, would probably be incomplete. Therefore, the user must compare the modeling published on the cloud of its project, with the latest default modeling. You may use the Cloud Connect tool for this analysis.

     

    1. ETL process update

    With the aid of Release Notes (available on the Customer Portal), the user may update the graphs of its project, in order to not interfere with its customizations. By completing the graphs development, as wishes, the user may re-deploy the project, according to the steps of the item ETL process update of Scenario 1.

    Warning

    Remember, use an own project for customizations and updates, and never edit the default project of TOTVS Smart Analytics.

     

    1. Update of Agent configuration files

    2. my.properties:

    To update my. properties, the user may also need to pay attention. Execute a comparison between the my.properties published on the Customer Portal and my.properties customized, to perform the proper alterations. In this case, as suggestion, you may use the text comparison program apdiff.

    Warning

    The project ID information, Smart Analytics login, access to Database and path for JDBC may be copied from the previous version, if not edited.

    1. run.bat

    To check this file, follow the same steps of the Scenario 1, on item “run.bat”.

     

    1. Conclusion

    Finally, the update cycle is completed, and from then on, the Schedule of load execution may be executed again.

     

    Notes:

    Release related:

    Flow: Smart Analytics Customization