Home
Accelerator Plus Administrator Guide
This guide is written for the Accelerator Plus system administrator who needs to configure and manage one or more Accelerator Plus instances.
Frequently Asked Questions and Troubleshooting Tips
This section provides recommendations to obtain the maximum performance from Accelerator Plus.
Altair Accelerator Plus Agent Preemption

Welcome
What's New
View new features for Accelerator Plus 2025.1.2.
Accelerator Plus Overview
Accelerator Plus is a high-performance hierarchical scheduler designed for distributed High Performance Computing (HPC) environments.
Accelerator Plus Tutorials
Currently there are no Accelerator Plus tutorials available.
Accelerator Plus User Guide
Accelerator Plus is a high-performance hierarchical scheduler designed for distributed High Performance Computing (HPC) environments.
Accelerator Plus Administrator Guide
This guide is written for the Accelerator Plus system administrator who needs to configure and manage one or more Accelerator Plus instances.
- System Overview
  Accelerator Plus is a high-performance hierarchical scheduler designed for distributed High Performance Computing (HPC) environments.
- Accelerator Plus Server and System Configuration
- Log Files
- Jobclasses
- Resource Management
  Altair Accelerator includes a subsystem for managing computing resources. This allows the design team to factor in various constraints regarding hardware and software resources, as well as site policy constraints.
- Frequently Asked Questions and Troubleshooting Tips
  This section provides recommendations to obtain the maximum performance from Accelerator Plus.
  - Accelerator Plus Modulation
  - Altair Accelerator Plus Agent Preemption

Altair Accelerator Plus Agent Preemption

Preempting Accelerator Plus Demand Jobs on a Base Queue

An administrator may need to preempt an Accelerator Plus demand job from the underlying Accelerator base queue. This may be useful in the case that other priority work is more important than a long-running Accelerator Plus set of jobs in a scheduling bucket that is hosted by a long-running demand job on a base queue.

To perform the preemption of the demand job, send it the SIGUSR1 signal as follows:

nc -q $BASEQ preempt -method "0:*:SIGUSR1" 159500

When the demand job receives SIGUSR1, it will cease accepting jobs to run and then terminate the running Accelerator Plus job, transitioning it to WITHDRAWN state. Then the demand job exits with "Done" status. The Accelerator Plus job will be transitioned to QUEUED state and will be scheduled by Accelerator Plus to be rerun.