Differences
This shows you the differences between two versions of the page.
| Next revision | Previous revision | ||
|
compute_slurm_reservations [2019/06/17 13:50] ktm5j created |
compute_slurm_reservations [2021/10/08 17:28] (current) |
||
|---|---|---|---|
| Line 1: | Line 1: | ||
| - | ====== SLURM Reservations ====== | + | ==== SLURM Reservations ==== |
| - | ===== Getting a Reservation ===== | + | === Getting a Reservation === |
| - | Reservations for slurm nodes/resources can be requested by submitting a ticket by email to <cshelpdesk@virginia.edu>. We can give out reservations for a specific amount of computing resources (Cores, Mem, GPU count, etc.) or by reserving an entire compute node or node group (i.e. slurm1 or slurm[1-5]). | + | Reservations for slurm nodes/resources can be requested by submitting a ticket by email to <cshelpdesk@virginia.edu>. We can give out reservations for a specific amount of computing resources (Cores, Mem, GPU count, etc.) or by reserving an entire compute node or node group (ex. "cortado01" or "cortado[01-05]"). |
| In your request for a reservation, please be specific about: | In your request for a reservation, please be specific about: | ||
| - | - What resources are needed | + | - What servers are needed (ex. "cortado01-05") |
| - | - How long they will be needed | + | - The start and end time of the reservation (ex. today to Dec 31st) |
| - | - Who has rights to use reservation (list of user IDs) | + | - Who can use the reservation (a list of user IDs) |
| - | ===== Using Reservations ===== | + | //Note: the node must be free of any running jobs before a reservation can be created. So if you are requesting a reservation on a node upon which you are running a job, we can't create the reservation until the job completes on the node. So it is best to cancel ('scancel') the job before requesting a reservation.// |
| + | |||
| + | //Note: nodes may be reserved for a maximum of fourteen days, and extended in one week increments.// | ||
| + | |||
| + | === Using reservations when submitting jobs === | ||
| + | |||
| + | If you have been granted a reservation you will receive a reservation tag, a short string that is required to submit jobs. This string must be included in any ''%%srun%%'' or ''%%sbatch%%'' submissions via the flag ''%%--reservation=...%%''. | ||
| + | |||
| + | If our tag is "abc1de_4" then you would submit jobs using the flag ''%%--reservation=abc1de_4%%'' | ||
| + | |||
| + | This can be done at the command line: | ||
| + | |||
| + | <code> | ||
| + | [abc1de@portal03 ~]$ sbatch --reservation=abc1de_4 slurm.sh | ||
| + | ... | ||
| + | [abc1de@portal03 ~]$ srun --reservation=abc1de_4 slurm.sh | ||
| + | </code> | ||
| + | |||
| + | Or you can include the flag in the header of your sbatch file: | ||
| + | |||
| + | <code> | ||
| + | [abc1de@portal03 ~]$ cat reservation.sh | ||
| + | #!/bin/bash | ||
| + | #SBATCH --job-name="Reservation Example" | ||
| + | # | ||
| + | #SBATCH --error="stderr.txt" | ||
| + | #SBATCH --output="stdout.txt" | ||
| + | # | ||
| + | #SBATCH --reservation=abc1de_4 <- Include reservation tag | ||
| + | ... | ||
| + | [abc1de@portal03 ~]$ sbatch reservation.sh | ||
| + | </code> | ||
| + | |||
| + | === Listing Reservations === | ||
| + | |||
| + | You can see a listing of all active reservations by using the ''%%scontrol%%'' command: | ||
| + | |||
| + | <code> | ||
| + | [abc1de@portal03 ~]$ scontrol show res | ||
| + | ReservationName=abc1de_4 StartTime=2019-06-25T00:00:00 EndTime=2019-07-02T16:00:00 | ||
| + | Nodes=slurm1 NodeCnt=1 CoreCnt=12 Features=(null) PartitionName=(null) Flags=SPEC_NODES | ||
| + | TRES=cpu=24 | ||
| + | Users=abc1de Accounts=(null) Licenses=(null) State=ACTIVE BurstBuffer=(null) Watts=n/a | ||
| + | |||
| + | </code> | ||
| - | If you have been granted a reservation you will receive a reservation tag, a short string that is required to submit jobs. This string must be included in any ''%%srun%%'' or ''%%sbatch%%'' submissions via the flag ''%%--reservation=tagstr_3%%'' | ||