10
Scheduling systems. SLURM example. Lauri Anton

Scheduling systems. SLURM example. - ut

  • Upload
    others

  • View
    6

  • Download
    1

Embed Size (px)

Citation preview

Page 1: Scheduling systems. SLURM example. - ut

Scheduling systems.SLURM example.

Lauri Anton

Page 2: Scheduling systems. SLURM example. - ut

11.06.2009 HPC@UT 2

Example of a resource (old)

• Juhtmasinad

• Arvutussõlmed

• Ethernet

• InfiniBand

• Jahutid

• Gaaskustutus

Pildilt väljas:

• UPS

• generaator

Page 3: Scheduling systems. SLURM example. - ut

Example of a resource (new)

Page 4: Scheduling systems. SLURM example. - ut

Why scheduling system?

• Resources are limited• CPU cores/cycles

• RAM

• Network bandwith

• Scratch space (/tmp)

• Many users with different needs

• Jobs need to be ordered

• Priorities

• Protecting users from themselves

• Protecting jobs from other jobs

Page 5: Scheduling systems. SLURM example. - ut

Ordering jobs

• What jobs should be run first?

• Parallel jobs vs single-core

Page 6: Scheduling systems. SLURM example. - ut

Historical overview of schedulers

• Pre-scheduler times

• PBS - Portable Batch SystemTorque + Maui

• Various grid middlewares

• SLURM

Page 7: Scheduling systems. SLURM example. - ut

SLURM overview

Page 8: Scheduling systems. SLURM example. - ut

Partitions, jobs, job steps

Page 9: Scheduling systems. SLURM example. - ut

Aims of the scheduling policy in HPC@UT

• Long jobs have to be limited

• Short jobs should start quickly

• Parallel jobs should start quickly

• Large users should have lower priority (fairshare)

Page 10: Scheduling systems. SLURM example. - ut

Home reading

• https://computing.llnl.gov/linux/slurm/quickstart.html