Lifecycle Query Engine - Relational Store (LQE RS)

Authors: ChaoWang, KotTontranakwong
Build basis: Lifecycle Query Engine 7.0.3+, 7.x

This page provides information for those who are looking to install or migrate to Lifecycle Query Engine - Relational store (LQE RS) application. LQE rs is an implementation of the Lifecycle Query Engine that replaces Jena with a relational database. The information in this page can help you determine the topology that fits your requirements.

Setup LQE RS on an Existing LQE server

When to use:

• The LQE Jena server does not have much data and is not under heavy load.

Pros:

• No need to setup a separate LQE instances on a new server.
• Less JRS data sources to manage.
• Less risk when transitioning to LQE RS.
• LQE Jena can be decommissioned after transitioning to LQE RS. (Need to find documentation for decommissioning)

Cons:

• There could be a need to move the server off the powerful machine in the future (after LQE Jena is completely decommissioned). This raises a need for a server rename.

Setup LQE RS on a New/Separate LQE server

When to use:

• The current LQE has a large index and is under heavy load.
• You do not have LQE in the ELM environment and want to add LQE RS to the environment.

Pros:

• LQE RS does not require as much resources as LQE Jena; therefore, can be installed on a smaller machine.
• LQE RS will not be impacted by the LQE Jena’s performance.
• There could be a need to move the server off the powerful machine in the future (after LQE Jena is completely decommissioned). This can avoid the need for a server rename.

Cons:

• More data sources in JRS to manage

Use a New or Existing data warehouse database for LQE RS?

Use the following links to help in determining whether to use the existing DW database or create a new DW database for LQE RS.

• https://jazz.net/wiki/bin/view/Deployment/LQERSDatabaseSizing

When setting LQE RS with a new DW database, ensure the DW database is not empty.

• You will need to temporarily change DW connection in JTS to point to the new database, use JTS setup to populate the new DW database, then, switch the DW connection in JTS back to the original DW.

When using the same DW, if you need to disable access control, it will be done by unchecking the checkbox in the “Data Warehouse” data source in the Report Builder. This will disable access control for both data warehouse and LQE RS.

Steps to setup LQE RS

1. (NEW LQE only) Install LQE – follow the steps to install and setup LQE
2. Ensure LQE RS connect to the desired DW database:
   1. Go to LQE > Administration > Relational Store > Settings at https://server:port/lqe/web/pgs/settings
   2. Ensure the database connection information under the “Data Warehouse Connection” section is correct and that the connection is successful.
      
      
      
3. Disable/Enable Relational Store:
   1. Go to LQE > Administration > Relational Store > Settings at https://server:port/lqe/web/pgs/settings
   2. Under the “Relational Store Settings” section, check or uncheck the “Enabled” setting
      
      
      When switching from LQE Jena to LQE RS, make sure to reindex the vocabularies.
      
      
4. Disable/Enable TRS processing for an LQE node:
   1. Go to LQE > Administration > Nodes at https://server:port/lqe/web/admin/nodes
   2. Select a node you want to configure by clicking the Node name in the column header
   3. Check/Uncheck the TRS Processing setting as needed.
      
      
      
5. Allow/Disallow SPARQL Service on LQE:
   1. Go to LQE > Administration > SPARQL Service at https://server:port/lqe/web/admin/query
   2. Check/Uncheck the “Query Service Enabled” setting
      
      
      

Data Sources in Report Builder (JRS)

In JRS, when you discover and connect data sources, it looks for LQE that’s registered with the JTS and create the following data sources for each LQE:

• Lifecycle Query Engine (/sparql)
• Lifecycle Query Engine scoped by a configuration (/sparql-variable-config)
• Data Warehouse

Assuming you have 2 LQEs and 1 DW database, JRS will detect 5 data sources

• 2 Lifecycle Query Engine ==> one for each LQE
• 2 Lifecycle Query Engine scoped by a configuration ==> one for each LQE
• Data Warehouse
  
  
  

And if you have 2 LQEs with 2 DW databases, JRS will detect 6 data sources; one additional Data Warehouse data source that LQE RS is connected to.

What to consider when managing data sources in Report Builder?

• Existing reports are tied to the data source by data source ID.
  • For existing SPARQL reports to use LQE RS (SQL), the data source that is used by the reports must be configured to connect to LQE RS’s data warehouse.
  • Benefits:
    • This allows you to switch between SPARQL and SQL in the existing reports.
    • You can compare the reports’ results between SPARQL and SQL and ensure the results are expected.
    • This is used to migrate existing reports to LQE RS (SQL)
  • A data source cannot be deleted if there is one or more reports using it.
• Archive un-use data sources so that they are not visible to the users when creating or editing a report.
  • Archive is a better option than Delete as it ensures the data source is available for any back-end usage and that no new data source is detected if there is a need to reload the data sources in JRS again.
    
    

Configuring data sources in Report Builder

Before you begin

• Review the “What to consider when managing data sources in Report Builder” section and determine which data sources to use.
• Archive the data sources that you do not want users to create reports against.

Steps to enable Relational Store in JRS data sources:

1. In JRS > Administration > Data sources page, select the “Lifecycle Query Engine” data source by clicking the data source name
2. Under Additional Properties section, select ‘Data Warehouse’ from the dropdown
   
   
3. Repeat step #1 and #2 for “Lifecycle Query Engine scoped by a configuration” data source. Select “Data Warehouse (indexed by: Lifecycle Query Engine)” as Lifecycle Query Engine relational store instead of “Data Warehouse” in step #2.

Converting existing reports from SPARQL to SQL:

• The query is automatically changed when switching from SPARQL to SQL
• The only exception is reports with custom expressions. Manual modifications are required.
  
  

Q&A

Is it possible to flip all existing reports from SPARQL to SQL?

• Yes, you can add the following property to app.properties

  lqe.dw.trs.override = true

• A report is converted at run-time (same as before) and does not impose any additional burden on the Report Builder server. The setting can be removed at anytime and it does not impact any existing reports.

????? Does it mean that if I switch from SPARQL to SQL that the application will try to convert those custom reports? or is it going to skip them completely?

• It will not try to convert them. Reports will not run and there will be an error message.

How to configure Report Builder to use LQE RS (SQL) by default?

• https://www.ibm.com/docs/en/engineering-lifecycle-management-suite/lifecycle-management/7.1?topic=experience-advanced-properties
• The following setting can be set

  lqe.dw.trs.default

Limitations

• https://www.ibm.com/docs/en/engineering-lifecycle-management-suite/lifecycle-management/7.1?topic=builder-limitations-lifecycle-query-engine-relational-store
• Server rename not supported in 703, but is in 7.1

Related topics: Deployment web home, Deployment web home

External links:

• IBM

Additional contributors: TWikiUser, TWikiUser