SLURM usage summary: Difference between revisions

From crtc.cs.odu.edu
Jump to navigation Jump to search
Ctsolakis (talk | contribs)
BrainMapping, CNF, HPC, I2MConversion, NDAM, Segmentation, UltraSound3D, Bureaucrats, external, internal, Administrators
2,152 edits
Created page with "= View en-queued jobs = squeue -u <username> = View available nodes = sinfo --state=idle"
 
Ctsolakis (talk | contribs)
BrainMapping, CNF, HPC, I2MConversion, NDAM, Segmentation, UltraSound3D, Bureaucrats, external, internal, Administrators
2,152 edits
No edit summary
 
(7 intermediate revisions by the same user not shown)
Line 6: Line 6:


  sinfo --state=idle
  sinfo --state=idle
= Create and interactive jobs =
salloc [SBATCH ARGUMENTS]
= SBATCH Arguments =
* <code> -c <number of cores> </code>  OR <code> --cpus-per-task </code> e.g <code> -c 12 </code> for allocating 12 cores
* <code> -N <number of nodes> </code> OR <code> --nodes <number of nodes> </code> e.g <code> -c 4 </code> for allocating 4 nodes
* <code> -p <name> </code> OR <code> --partition <name> </code> e.g <code> --partition "himem" </code> for using a node of the high memory group
* <code> -C <name> </code> OR <code> --constrain <name> </code>  e.g <code> --constrain "coreV2*" </code> for using only nodes with hostname starting with coreV2*
* <code> --exclusive </code> block other jobs running in the node(s)
= SLURM Cheetsheet =
https://docs.hpc.odu.edu/#slurm-cheat-sheet
= HPC ODU Documentation =
https://docs.hpc.odu.edu/
= Turing cluster =
** TODO add GPU information **
{| class="wikitable"
! hostname
! nodes
! memory (GB)
! cache (MB)
! Model Name
! Turbo
! CPUs per node
! Threads per core
! Cores per socket
! sockets
|-
| coreV1-22-0*
| 28
| 127
| 20
| Intel(R) Xeon(R) CPU E5-2660    @ 2.20GHz
| 3.00GHz
| 16
| 1
| 8
| 2
|-
| coreV2-22-0*
| 36
| 126
| 25
| Intel(R) Xeon(R) CPU E5-2660 v2 @ 2.20GHz
| 3.00GHz
| 20
| 1
| 10
| 2
|-
| coreV2-25-0*
| 76
| 126
| 25
| Intel(R) Xeon(R) CPU E5-2670 v2 @ 2.50GHz
| 3.30GHz
| 20
| 1
| 10
| 2
|-
| coreV2-23-himem-*
| 4
| 757
| 16
| Intel(R) Xeon(R) CPU E5-4610 v2 @ 2.30GHz
| 2.70GHz
| 32
| 1
| 8
| 4
|-
| coreV3-23-0*
| 50
| 125
| 40
| Intel(R) Xeon(R) CPU E5-2698 v3 @ 2.30GHz
| 3.20GHz
| 32
| 1
| 16
| 2
|-
| coreV4-21-0*
| 30
| 125
| 40
| Intel(R) Xeon(R) CPU E5-2683 v4 @ 2.10GHz
| 3.00GHz
| 32
| 1
| 16
| 2
|-
| coreV4-21-himem-*
| 3
| 504
| 40
| Intel(R) Xeon(R) CPU E5-2683 v4 @ 2.10GHz
| 3.00GHz
| 32
| 1
| 16
| 2
|}
= Wahab cluster =
{| class="wikitable"
! hostname
! nodes
! memory (GB)
! cache (MB)
! Model Name
! Turbo
! CPUs per node
! Threads per core
! Cores per socket
! sockets
|-
| any
| 158
| 384
| -
| Intel® Xeon® Gold 6148    @ 2.40GHz
| 3.70GHz
| 40
| 1
| 20
| 2
|}

Latest revision as of 11:57, 12 June 2019

View en-queued jobs

squeue -u <username>

View available nodes

sinfo --state=idle

Create and interactive jobs

salloc [SBATCH ARGUMENTS]

SBATCH Arguments

  • -c <number of cores> OR --cpus-per-task e.g -c 12 for allocating 12 cores
  • -N <number of nodes> OR --nodes <number of nodes> e.g -c 4 for allocating 4 nodes
  • -p <name> OR --partition <name> e.g --partition "himem" for using a node of the high memory group
  • -C <name> OR --constrain <name> e.g --constrain "coreV2*" for using only nodes with hostname starting with coreV2*
  • --exclusive block other jobs running in the node(s)

SLURM Cheetsheet

https://docs.hpc.odu.edu/#slurm-cheat-sheet

HPC ODU Documentation

https://docs.hpc.odu.edu/

Turing cluster

    • TODO add GPU information **
hostname nodes memory (GB) cache (MB) Model Name Turbo CPUs per node Threads per core Cores per socket sockets
coreV1-22-0* 28 127 20 Intel(R) Xeon(R) CPU E5-2660 @ 2.20GHz 3.00GHz 16 1 8 2
coreV2-22-0* 36 126 25 Intel(R) Xeon(R) CPU E5-2660 v2 @ 2.20GHz 3.00GHz 20 1 10 2
coreV2-25-0* 76 126 25 Intel(R) Xeon(R) CPU E5-2670 v2 @ 2.50GHz 3.30GHz 20 1 10 2
coreV2-23-himem-* 4 757 16 Intel(R) Xeon(R) CPU E5-4610 v2 @ 2.30GHz 2.70GHz 32 1 8 4
coreV3-23-0* 50 125 40 Intel(R) Xeon(R) CPU E5-2698 v3 @ 2.30GHz 3.20GHz 32 1 16 2
coreV4-21-0* 30 125 40 Intel(R) Xeon(R) CPU E5-2683 v4 @ 2.10GHz 3.00GHz 32 1 16 2
coreV4-21-himem-* 3 504 40 Intel(R) Xeon(R) CPU E5-2683 v4 @ 2.10GHz 3.00GHz 32 1 16 2

Wahab cluster

hostname nodes memory (GB) cache (MB) Model Name Turbo CPUs per node Threads per core Cores per socket sockets
any 158 384 - Intel® Xeon® Gold 6148 @ 2.40GHz 3.70GHz 40 1 20 2
Retrieved from "https://crtc.cs.odu.edu/index.php?title=SLURM_usage_summary&oldid=2908"