Understanding the details of the Spotfire Statistica Monitoring and Alerting Server (a.k.a MAS) log files

Understanding the details of the Spotfire Statistica Monitoring and Alerting Server (a.k.a MAS) log files

book

Article ID: KB0137928

calendar_today

Updated On:

Products Versions
Spotfire Statistica 14.x
Spotfire Data Science - Statistica 14.x

Description

Below is an example of some of the information a MAS log contains:

INFO     2025-08-23 07:12:02.104  8252  9096  SVC: Start running taskset TASKSET 1
INFO     2025-08-23 07:12:02.104  8252  8256  ROS initialization beginning for monitor TASKSET 1 in taskset TASKSET 1
WARNING  2025-08-23 07:13:02.108  8252  8256  Error 0x8004E024, [Activation Timeout] while trying to create the Running Object to run monitor 'TASKSET 1' in taskset 'TASKSET 1'. Attempt number 1. Trying again.
INFO     2025-08-23 07:14:02.760  8252  6632  ROS initialization succeeded for monitor TASKSET 1 in taskset TASKSET 1. IPooledObj in dllhost process 10432.
INFO     2025-08-23 07:15:35.857  8252  6632  SVC: Start running monitor TASKSET 1 (revision 0) in taskset TASKSET 1
INFO     2025-08-23 07:15:36.29   11804 2892  Runner: Running Monitor TASKSET 1 in Taskset TASKSET 1
INFO     2025-08-23 07:15:43.259  11804 2892  Runner: Finished running Monitor TASKSET 1 in Taskset TASKSET 1
INFO     2025-08-23 07:15:50.76   8252  6632  SVC: finished running monitor TASKSET 1 in taskset TASKSET 1, 0 new alarms recorded
INFO     2025-08-23 07:15:50.86   8252  6632  Task [TASKSET 1] in Taskset [TASKSET 1] Succeeded; Times: [Queued]: 0.000000 [COM+ Object Creation]: 120.344000 [COM+ Object Initialization]: 0.313000 [Pre-execution]: 93.093000 [Executing]: 7.563000 [Recording Alarms]: 6.640000 [Total]: 227.953000
INFO     2025-08-23 07:17:44.265  8252  6632  Taskset [TASKSET 1] Finished; Times: [Running Tasks]: 227.985000 [Processing Rules]: 0.015000 [Total]: 228.000000

Environment

OS:  Windows

Resolution

1. The Execution Lifecycle

When a taskset is triggered, the MAS log records events in a specific sequence. Understanding this flow is essential for identifying where a process is failing.

StepLog Message / EventDescription
1SVC: Start running tasksetThe MAS service begins processing the taskset.
2ROS initialization beginningAttempting to initialize the Running Object (ROS).
3Error 0x8004E024Warning: Activation Timeout. MAS will retry 2 more times (3 total) before failing.
4ROS initialization succeededThe dllhost process successfully creates the MASTaskRunner.
5SVC: Start running monitorService hand-off to the specific monitor/analysis begins.
6Runner: Running MonitorThe logic within MASTaskRunner is now executing.
7Runner: Finished running MonitorTask execution logic is complete.
8Task [Name] Succeeded/FailedSummary of specific task results and timing stats.
9Taskset [Name] FinishedFinal summary of the entire taskset execution.

Note on Concurrent Execution: If you see a warning stating “Taskset is still running and cannot be executed,” it means the previous instance hasn't finished. MAS will automatically reschedule the next run.


2. Interpreting Performance Statistics

At the end of a task, MAS provides a breakdown of time spent (in seconds). Use these to find the bottleneck:

  • [Queued]: Time spent in the MAS service thread pool before being picked up. (Separate from COM+ pool).

  • [COM+ Object Creation]: Duration to create the COM+ application object.

  • [COM+ Object Initialization]: Duration to initialize the created COM+ object.

  • [Pre-execution]: The gap between COM+ initialization and the start of actual execution.

  • [Executing]: The actual runtime of the task logic (MASTaskRunner.ExecuteRunningObj).

  • [Recording Alarms]: Time spent writing alarm data into the Meta Database.

  • [Total]: The sum of all metrics above.


3. Advanced Configuration & Tuning

Managing High Taskset Volumes (Cache Optimization)

In environments with many tasksets, the MAS service can spend 45+ minutes reading the database to update its cache, which can lag the system during heavy execution windows (e.g., 2:00 AM).

  • Recommendation: Increase the SEWSSPOLLING value in MASConfig.xml.

  • Values: Setting this to 1800 (30 mins) or 3600 (1 hour) prevents the service from interrupting task execution to refresh the cache.

Understanding COM+ Limits and Timeouts

The interaction between MAS and COM+ is governed by rules in the COM+ settings.  See https://support.tibco.com/support-home/kbsearch/article?articleNumber=KB0135550

  • Activation Limit: Must be > 0. It is recommended to use a large number to prevent COM+ from constantly recycling instances, which causes significant overhead.

  • Expiration Timeout: Critical: In COM+, this clock starts the moment an object is activated, not when it finishes.

  • The Conflict: If your Activation Limit is 1, and your Expiration Timeout is 15 minutes, your task will be killed at 15 minutes—even if the "Maximum time for tasks to run" in Enterprise Manager is set to 1 hour.

  • Best Practice: Set the Expiration Timeout to at least 30–60 minutes to ensure it doesn't override your MAS taskset limits.


4. Troubleshooting FAQ

Q: Why did my log stop updating?

A: This happens if the MASMonitorService was stopped. Restart the MASMonitorService service again to re-initialize the log hooks.

Q: What is the difference between "Acquiring," "Active," and "Executing"?

  • Queued: Tasks waiting for a free thread.

  • Acquiring Object: Threads currently attempting to fetch a COM+ instance.

  • Active Objects: Total pooled COM+ objects in existence (Queued + Acquiring + Executing).

  • Executing: Tasks actively running logic.

Q: Is there a limit on queued objects?

A: Yes. By default, the MAS thread pool size is matched to the COM+ pool size.

Issue/Introduction

Understanding the details of the Spotfire Statistica Monitoring and Alerting Server log files

Additional Information

What are the log levels for the Statistica Monitoring and Alerting Service (MAS)?

https://support.tibco.com/support-home/kbsearch/article?articleNumber=KB0072367

How to increase the Statistica Monitoring and Alerting Server Pool Size

https://support.tibco.com/support-home/kbsearch/article?articleNumber=KB0135550

How to troubleshoot the Spotfire Statistica Monitoring and Alerting server issues

https://support.tibco.com/support-home/kbsearch/article?articleNumber=KB0137867