Troubleshooting Analysis, Models, and Reports

For workspace artifacts (analysis, models, and reports) to function properly, you need to primarily ensure that there is a stable data flow, and a robust web and scheduler setup. Most of the issues occur due to incorrect configuration (reports, models, analysis are too large to execute) or lack of  data load to run the artifacts. When working with workspace analysis, models and reports, you may come across issues where the error is reported in the UI or in the logs.

Use this information in this topic to identify the symptoms and use the resolution to resolve the issue. For further assistance, create a BMC Support case.


Related topics

Analyzing Domain Data

Modeling Capacity Usage

Creating Reports and Report Templates

Viewing log information

To view logs from the command line: 

  1. Log on to the Application Server running the datahub.
  2. Navigate to $CPITBASE/web/log and check cpit.log* logs. 

You can also use the logGrabber utility to collect these logs. For details on logGrabber, see the knowledge article number 000244652 (Support logon ID required).

To view Golden (scheduled) Analysis/Models logs from the UI:

  1. Navigate to Administration > ETL & System Tasks > System Tasks. 
  2. Select the view logs button corresponding to the task of type "Golden Producer". Scheduled works are executed by Daily, Monthly or Weekly Reporter tasks according to the selected scheduling.
  3. From the logs page, you can download the logs using Collect logs option.

To view Report logs from the UI:

  1. Navigate to Administration > ETL & System Tasks > System Tasks. 
  2. Select the view logs button corresponding to the task of type reporter:
    •  On demand tasks are executed by "On demand Reports" task. 
    • Scheduled tasks are executed by Daily, Monthly, or Weekly Reporter tasks according to the selected scheduling.
  3. From the logs page, you can download the logs using Collect logs option.

Exporting and importing analysis and forecasting models

For details about exporting and importing analysis and forecasting models, see the following Knowledge Articles:

  • Export and Import Analyses Data and Settings
  • Export and Import Models Data and Settings
  • Export Data from a Workspace Domain - Domain extractor

Common issues

Issue

Cause

Resolution

Report is not received via email. The following message is reported in the log:

BCO_SCH_ERR011: Error sending mail
BCO_SCH_ERR016: Error sending mail

SMTP is unreachable or it is incorrectly configured.

The scheduled reports are not received via email or the Golden analysis or models are not generated.

The following message is reported in the log: $CPITBASE/scheduler/log/scheduler.out or scheduler.err

ERROR FATAL - Program abnormally terminated [java.lang.OutOfMemoryError: GC overhead limit exceeded]

Scheduler Java heap size insufficient

Increase the ETL Java heap size. For details, see KA: How to Enlarge Scheduler's Heap.

The report received via email is empty.

No data for the report to run.

  • If it's a basic report, check whether the analysis/models inside the report runs correctly.
  • Check whether the systems in the report have data for the metric and the time period selected in the scenario.
  • Check for ETL Data Load or Data Warehousing errors.

For details, see How to quickly assess the health of TrueSight Capacity Optimization.

After running a model, no output is generated. The following message is reported in the UI: 

No data available

Data set size (number of samples) is insufficient for the model.

  • Check that the entity filter is not empty.
  • Check the selected time filter period.
  • Check the systems in the model have data for the metric, resolution and the time period selected into the scenario, trying to change it to a larger one.

  • Check for ETL Data Load or Data Warehousing errors.

    For details, see KA: How to quickly assess the health of TrueSight Capacity Optimization.


When you execute a Performance versus Time Analysis PTA, no data is available. The following message is reported in the UI: 

No data available

Data set size (number of samples) is insufficient  for the PTA to run.

When you execute a Performance versus Load Analysis PLA. no data is available. The following message is reported in the UI: 

No data available

Data set size (number of samples) is insufficient (less than 30 points).

 

Tip: For faster searching, add an asterisk to the end of your partial query. Example: cert*