/
Why Pulse can’t always cancel TM1 threads

Why Pulse can’t always cancel TM1 threads

 

TM1 cancellation is cooperative: Pulse or TM1Top requests cancellation, IBM Planning Analytics (TM1) sets a cancel flag on the target thread, and the thread checks that flag only at safe cancellation points (between internal TM1 API operations). This protects data integrity and server stability.

  • Request comes from Pulse or TM1Top; TM1 sets a cancel flag.

  • Threads do not get killed at the OS level.

  • Thread exits only at safe points between internal API operations.

It is IBM Planning Analytics (TM1) responsibility to decide which thread can be cancelled or not.

Long-running internal operations keep the thread from checking the cancel flag until they finish. As a result, a thread can appear "cancelling" for minutes or hours depending on data volume, calculation complexity, and server load.

  • Request cancellation via Pulse or TM1Top and monitor progress.

  • If a thread remains non-responsive for too long, plan a controlled TM1 restart (last resort).

  • TM1: IBM Planning Analytics engine.

  • Pulse: Cubewise monitoring/administration tool.

  • TM1Top: IBM utility to view/cancel threads.

  • Add links to troubleshooting guides here.