A collection of CLI tools used on CARC HPC clusters for various Slurm tasks and queries.
The CLI tools are Python scripts that rely only on the Python standard library and use the argparse library. Some scripts may need to be modified to run on other HPC clusters, depending on the Python version available, Slurm configuration, etc. Check the notes in each script for more information.
To install, simply clone the repo:
git clone --depth 1 https://github.com/uschpc/slurm-tools.git
The scripts will be downloaded with execute permissions. If desired, move the scripts to another directory, such as a directory on PATH. If needed, install and/or load a compatible version of Python.
Each script contains help and usage information that can be viewed with the --help flag (e.g., jobinfo --help).
The scripts are described and shown with example output below.
View account information for user.
$ myaccount
-----------------------------------------------------------------
Project accounts
-----------------------------------------------------------------
User Account Cluster Default QOS
---------- --------------- ---------- --------------- -----------
ttrojan ttrojan_123 discovery ttrojan_123 normal
ttrojan ttrojan_123 condo ttrojan_123 normal
-----------------------------------------------------------------
Discovery account service units (SUs)
-----------------------------------------------------------------
Account Limit Usage Remaining
----------------- --------------- --------------- ---------------
ttrojan_123 12000000 422699 11577301
*** Usage based on resources reserved
*** 1 CPU/minute : 1 SU
*** 4 GB of memory/minute : 1 SU
*** 1 A100 or A40 GPU/minute : 8 SUs
*** 1 V100 or P100 GPU/minute : 4 SUs
*** 1 K40 GPU/minute : 2 SUs
-----------------------------------------------------------------
Allowed Endeavour partitions
-----------------------------------------------------------------
Partition Accounts
-------------- --------------------------------------------------
trojan ttrojan_123,hpcroot
shared ALL
View available resources on nodes.
$ noderes -p debug
-------------------------------------------------------------------
Node Partition State CPU GPU Free Free Free
Model Model CPUs Memory GPUs
------ ------------ --------- ----------- -------- ---- ------ ----
b11-09 debug idle epyc-7313 a40 32 248G 2
d05-41 debug idle xeon-4116 -- 24 185G --
d05-42 debug idle xeon-4116 -- 24 185G --
e23-02 debug mixed xeon-2640v4 p100 8 75G 2
View node information by partition, CPU/GPU models, and state.
$ nodeinfo
-----------------------------------------------------------------------------------------------------------------------
Partition Timelimit CPU model CPUs/ Memory(MB)/ GPU model State Nodes Nodelist
node node
------------ ----------- ------------- ----- ----------- ------------------ ------ ----- ------------------------------
debug 1:00:00 xeon-2640v4 20 128000 gpu:p100:2(S:0-1) mix 1 e23-02
debug 1:00:00 xeon-2650v2 16 63400 (null) idle 4 e05-[42,76,78,80]
debug 1:00:00 xeon-2640v3 16 63400 gpu:k40:2(S:0-1) idle 1 e09-18
epyc-64 2-00:00:00 epyc-7513 64 256000 (null) mix 1 a01-03
epyc-64 2-00:00:00 epyc-7542 64 256000 (null) mix 8 b22-[01-08]
epyc-64 2-00:00:00 epyc-7513 64 256000 (null) alloc 9 a01-[02,04,08,11,14,18],b01-[0
epyc-64 2-00:00:00 epyc-7513 64 256000 (null) idle 51 a01-[05,07,09,12-13,16-17,19],
epyc-64 2-00:00:00 epyc-7542 64 256000 (null) idle 24 b22-[09-32]
main* 2-00:00:00 xeon-4116 24 192000 (null) drain 2 d05-39,d06-27
main* 2-00:00:00 xeon-2640v4 20 64000 (null) drain 1 d17-38
main* 2-00:00:00 xeon-4116 24 94000+ (null) mix 47 d05-[06-15,26-38,40-42],d06-[1
main* 2-00:00:00 xeon-2640v4 20 64000 (null) mix 1 d18-34
main* 2-00:00:00 xeon-2640v3 16 63400 (null) mix 1 e06-03
main* 2-00:00:00 xeon-2640v3 16 63400 gpu:k40:2(S:0-1) mix 1 e07-12
main* 2-00:00:00 xeon-4116 24 94000 (null) alloc 3 d11-[19,30,40]
main* 2-00:00:00 xeon-2640v3 16 63400 (null) alloc 4 e13-[30-33]
main* 2-00:00:00 xeon-4116 24 94000 (null) idle 28 d11-[09-11,13-18,20-24,26-29,3
main* 2-00:00:00 xeon-2640v4 20 63400+ (null) idle 70 d17-[03-37],d18-[01-33,35],e16
main* 2-00:00:00 xeon-2640v3 16 63400 (null) idle 76 e06-[01-02,04,06-22],e11-[26-2
main* 2-00:00:00 xeon-2640v3 16 63400 gpu:k40:2(S:0-1) idle 16 e07-[01-11,13-16,18]
main* 2-00:00:00 xeon-2640v4 20 63400 gpu:k40:2(S:0-1) idle 40 e16-[01-04,06-24],e17-[01-04,0
gpu 2-00:00:00 epyc-7513 64 256000 gpu:a100:2(S:0-7) mix 7 a01-[01,06,15,20],b01-[01,06],
gpu 2-00:00:00 epyc-7282 32 256000 gpu:a40:2(S:0-1) mix 7 a02-[01,06,15],a03-[06,15],a04
gpu 2-00:00:00 xeon-6130 32 191000 gpu:v100:2(S:0-1) mix 28 d11-[02-04],d13-[02-11],d14-[0
gpu 2-00:00:00 xeon-2640v4 20 128000 gpu:p100:2(S:0-1) mix 23 d23-[10,13-16],e21-[02-13,16],
gpu 2-00:00:00 xeon-6130 32 191000 gpu:v100:2(S:0-1) alloc 1 d14-07
gpu 2-00:00:00 xeon-2640v4 20 128000 gpu:p100:2(S:0-1) alloc 1 e23-01
gpu 2-00:00:00 epyc-7282 32 256000 gpu:a40:2(S:0-1) idle 5 a02-20,a03-[01,20],a04-[01,20]
gpu 2-00:00:00 epyc-7513 64 256000 gpu:a100:2(S:0-7) idle 5 b01-[15,20],b02-[06,15,20]
gpu 2-00:00:00 xeon-2640v4 20 128000 gpu:p100:2(S:0-1) idle 14 e21-[01,14-15],e22-[02-12]
oneweek 7-00:00:00 xeon-2650v2 16 128000 (null) drain 4 e02-[40-41,43,45]
oneweek 7-00:00:00 xeon-2650v2 16 128000+ (null) mix 4 e02-[68-71]
oneweek 7-00:00:00 xeon-2650v2 16 128000+ (null) alloc 21 e01-[46,48,52,60],e02-[48-55,7
oneweek 7-00:00:00 xeon-2640v3 16 63400 (null) alloc 1 e06-24
oneweek 7-00:00:00 xeon-2650v2 16 128000+ (null) idle 18 e01-[62,64,76],e02-[42,44,46,5
oneweek 7-00:00:00 xeon-2640v3 16 63400 (null) idle 1 e10-12
largemem 7-00:00:00 epyc-7513 64 1024000 (null) mix 2 a02-10,a04-10
largemem 7-00:00:00 epyc-7513 64 1024000 (null) idle 2 a01-10,a03-10
View job queue information.
$ jobqueue -p largemem
-----------------------------------------------------------------------------------------
Job ID User Job Name Partition State Elapsed Nodelist(Reason)
------------ ---------- ------------ ---------- -------- ----------- --------------------
7453378 jesp Run_do52bran largemem PENDING 0:00 (QOSMaxMemoryPerUser
7453379 jesp Run_do6b.job largemem PENDING 0:00 (QOSMaxMemoryPerUser
7473836 ttrojan sys/dashboar largemem RUNNING 2:42:28 a04-10
7449562 snet run_study_4x largemem RUNNING 2-23:17:10 a02-10
7453377 jesp Run_do51bran largemem RUNNING 2-17:02:07 a01-10
7470944 huy435 rfmixchr1 largemem RUNNING 21:18:51 a02-10,a04-10
View compact history of jobs.
$ jobhist -p largemem
----------------------------------------------------------------------------------------------------
Job ID Startdate User Job Name Partition State Elapsed Nodes CPUs Memory
------------- ---------- ---------- ------------ ---------- ---------- ----------- ----- ---- ------
14690860 2024-07-29 ttrojan sim.sl largemem RUNNING 3-08:19:19 1 32 128G
14734145 2024-07-31 jesp sfla largemem RUNNING 2-21:46:24 1 64 998G
14738354 2024-07-31 snet interactive largemem COMPLETED 06:56:19 1 16 400G
14741823 2024-07-31 huy435 model_fit1 largemem COMPLETED 07:04:19 1 64 248G
14741846 2024-07-31 huy435 model_fit2 largemem COMPLETED 08:10:59 1 64 248G
14741918 2024-08-01 snet feature.sl largemem FAILED 00:02:16 1 8 300G
View detailed job information.
$ jobinfo 483699
Job ID : 483699
Job name : simdebug.sl
User : ttrojan
Account : ttrojan_123
Working directory : /project/ttrojan_123/sim
Cluster : discovery
Partition : debug
Nodes : 2
Nodelist : e05-[42,76]
Tasks : 32
CPUs : 32
GPUs : 0
State : COMPLETED
Exit code : 0:0
Submit time : 2024-01-26T14:56:23
Start time : 2024-01-26T14:56:24
End time : 2024-01-26T14:57:32
Wait time : 00:00:01
Reserved walltime : 00:10:00
Used walltime : 00:01:08
Used CPU walltime : 00:36:16
Used CPU time : 00:35:18
CPU efficiency : 97.37%
% User (computation) : 96.35%
% System (I/O) : 3.65%
Reserved memory : 120G
Max memory used : 64.74G (estimate)
Memory efficiency : 53.95%
Max disk write : 1.04M
Max disk read : 14.42M