Protected
Slurm mock questions are available after login. Redirecting...
If you are not redirected, login.
Courses / Slurm / Study / Self-Test
Multiple-choice practice organized into fixed-size sets
25 total questions · 2 sets · 20 per set
How to Use
Set 1
Set 1 · Question 1 · Chapter 1 · SLM-01
Correct answer: To coordinate fair shared access to limited resources
Multi user environments need policy driven arbitration
Set 1 · Question 2 · Chapter 2 · SLM-02
Correct answer: slurmctld
slurmctld controls scheduling and cluster state
Set 1 · Question 3 · Chapter 3 · SLM-03
Correct answer: Generic resources such as GPUs and licenses
GRES extends scheduling beyond CPUs and memory
Set 1 · Question 4 · Chapter 4 · SLM-04
Correct answer: A queue abstraction with policy and node scope
Partitions define access limits and scheduling behavior
Set 1 · Question 5 · Chapter 5 · SLM-05
Correct answer: slurm.conf
slurm.conf is the central scheduler configuration file
Set 1 · Question 6 · Chapter 6 · SLM-06
Correct answer: Credential signing and trust authentication
Munge secures node to node communication
Set 1 · Question 7 · Chapter 7 · SLM-07
Correct answer: sbatch
sbatch is the batch submission interface
Set 1 · Question 8 · Chapter 8 · SLM-08
Correct answer: pending
Submitted jobs wait in queue before allocation
Set 1 · Question 9 · Chapter 9 · SLM-09
Correct answer: To run many similar tasks efficiently
Arrays scale repeated batch workloads
Set 1 · Question 10 · Chapter 10 · SLM-10
Correct answer: It improves utilization without delaying reserved priority work
Backfill fills resource gaps safely
Set 1 · Question 11 · Chapter 11 · SLM-11
Correct answer: age
Age prevents starvation for long waiting jobs
Set 1 · Question 12 · Chapter 12 · SLM-12
Correct answer: Limits caps boosts and preemption permissions
QoS provides multi tenant governance controls
Set 1 · Question 13 · Chapter 13 · SLM-13
Correct answer: afterok
afterok enforces success gated chaining
Set 1 · Question 14 · Chapter 14 · SLM-14
Correct answer: task cgroup
task cgroup applies CPU memory and device limits
Set 1 · Question 15 · Chapter 15 · SLM-15
Correct answer: Four GPUs for the job allocation
GRES syntax controls GPU reservation count
Set 1 · Question 16 · Chapter 16 · SLM-16
Correct answer: Components requiring different resource types
Heterogeneous jobs combine mixed allocations in one workflow
Set 1 · Question 17 · Chapter 17 · SLM-17
Correct answer: slurmdbd
slurmdbd centralizes accounting data
Set 1 · Question 18 · Chapter 18 · SLM-18
Correct answer: job wait time
Wait time reflects scheduling pressure and fairness
Set 1 · Question 19 · Chapter 19 · SLM-19
Correct answer: To fail over if the primary controller fails
Controller redundancy improves uptime
Set 1 · Question 20 · Chapter 20 · SLM-20
Correct answer: Cross cluster scheduling and sharing
Federation unifies multiple clusters operationally
Set 2
Set 2 · Question 1 · Chapter 21 · SLM-21
Correct answer: suspend and resume
Power lifecycle controls enable elastic efficiency
Set 2 · Question 2 · Chapter 22 · SLM-22
Correct answer: Configuration deviation from desired state
Drift control keeps clusters consistent
Set 2 · Question 3 · Chapter 23 · SLM-23
Correct answer: Apptainer
Apptainer and Singularity are common HPC runtimes
Set 2 · Question 4 · Chapter 24 · SLM-24
Correct answer: NCCL MPI CUDA on GPU nodes
This stack supports distributed accelerated training
Set 2 · Question 5 · Chapter 25 · SLM-25
Correct answer: Topology aware placement
Locality aware scheduling reduces communication overhead
Set 2 · Question 6 · Chapter 1 · SLM-01
Correct answer: To coordinate fair shared access to limited resources
Multi user environments need policy driven arbitration
Set 2 · Question 7 · Chapter 2 · SLM-02
Correct answer: slurmctld
slurmctld controls scheduling and cluster state
Set 2 · Question 8 · Chapter 3 · SLM-03
Correct answer: Generic resources such as GPUs and licenses
GRES extends scheduling beyond CPUs and memory
Set 2 · Question 9 · Chapter 4 · SLM-04
Correct answer: A queue abstraction with policy and node scope
Partitions define access limits and scheduling behavior
Set 2 · Question 10 · Chapter 5 · SLM-05
Correct answer: slurm.conf
slurm.conf is the central scheduler configuration file
Set 2 · Question 11 · Chapter 6 · SLM-06
Correct answer: Credential signing and trust authentication
Munge secures node to node communication
Set 2 · Question 12 · Chapter 7 · SLM-07
Correct answer: sbatch
sbatch is the batch submission interface
Set 2 · Question 13 · Chapter 8 · SLM-08
Correct answer: pending
Submitted jobs wait in queue before allocation
Set 2 · Question 14 · Chapter 9 · SLM-09
Correct answer: To run many similar tasks efficiently
Arrays scale repeated batch workloads
Set 2 · Question 15 · Chapter 10 · SLM-10
Correct answer: It improves utilization without delaying reserved priority work
Backfill fills resource gaps safely
Set 2 · Question 16 · Chapter 11 · SLM-11
Correct answer: age
Age prevents starvation for long waiting jobs
Set 2 · Question 17 · Chapter 12 · SLM-12
Correct answer: Limits caps boosts and preemption permissions
QoS provides multi tenant governance controls
Set 2 · Question 18 · Chapter 13 · SLM-13
Correct answer: afterok
afterok enforces success gated chaining
Set 2 · Question 19 · Chapter 14 · SLM-14
Correct answer: task cgroup
task cgroup applies CPU memory and device limits
Set 2 · Question 20 · Chapter 15 · SLM-15
Correct answer: Four GPUs for the job allocation
GRES syntax controls GPU reservation count