One of the things more useful when you are running a Pentaho production environment with a lot of users accessing the BI server, using reports, dashboards, OLAP analysis... is monitor the whole user performance.

That's why we've created STMonitoring. It includes a set of predefined dashboards, reports and olap analysis based on several monitoring models including:

User session events model:

• Analysis by year, month, day, hour and minute.
• Session events (login, logout).
• Session status (abandoned, ended, started).
• Session duration (avg, min, max).
• Session count by time (avg, max, concurrent).

Server content access model:

• Analysis by year, month, day, hour and minute.
• User.
• Content type:
  • Dashboards: STDashboard, STAgile, CDE.
  • Analysis: STPivot, Pentaho Analyzer, Saiku Analytics.
  • Reports: STReport, Pentaho Reporting.
• Content extension:
  • prpt
  • wcdf
  • xaction
  • xjpivot
• Content (complete path):
  • Duration.
  • Avg. duration.
  • Access count.

STAudit is a component developed by StrateBI for monitoring the status of ETL processes deployed in production systems.

We can extract the following information at the process and activity level (a process is composed of 1 or N activities):

• Processing time.
• Start and end time.
• Number of errors.
• Process status.

Although STAudit comes with predefined reports and dashboards, we can analyze the information collected in STAudit's database as needed.

With the predefined metadata the end user will be able to create reports using the STPivot, STDashboard and STReport tools.

In the ETL to be monitored, we will have to place marks in our ETL process. These marks will insert records in our STAudit database which we will later analyze in Pentaho.

In the server we will configure the connection with the DB, we will import metadata, reports and predefined dashboards.

STValidations is a LinceBI plugin that allows verifying the differences at data level between the source systems (transactional) and the DW after the execution of a loading process.

For this purpose, jobs have been implemented for both Pentaho Data Integration and Talend that allows you to compare the differences between the query in the source system and the target system (DW), in order to check the consistency of the data.

By using a table containing the query pairs to be compared, STValidations will automatically run each pair and compare the data provided by both queries, to later save the result of each comparison in another table.

The resources that we will be able to use are:

• OLAP metadata.
• Reporting metadata.
• Dashboards with Pentaho CDE.
• Dashboards with PowerBI.

Arquery's XMLA connector provides OLEDB data connectivity for OLAP (ODBO) clients and XMLA data sources. The connector is designed to interface with XMLA compliant data sources such as SAP, Mondrian, Jedox and other data sources.

The connector is ODBO-compatible and as such, it loads smoothly on ODBO clients. This allows Excel (for example) to be connected to the data sources listed above.

An obvious advantage of using this technology is to bring the powerful pivot table technology of Excel to XMLA stores. As such, the connector user has a basic BI tool at their fingertips with corresponding advantages in knowledge reuse and price.

During the design of this connector, the optimization of speed and memory consumption was taken into account. An MDX query can result in a large amount of data, and the system was designed with this in mind. As a result, the system is fast and robust, with low memory requirements.

The connector exposes the IMDDataset interface for classic connectivity. This interface allows axis retrieval and sequential access to data. In addition, mass data access is provided through the implementation of IMDRangeRowset. This allows for writing more effective and ultimately faster clients.