Default Resources¶
Cluster available resources¶
| Node type | Name(s) | #Threads | Memory | Processor(s) | GPU |
|---|---|---|---|---|---|
| standard (x38) | n001 to n038 |
128 | 2 TiB | 2 x AMD EPYC 7713 | - |
| bigmem | bigmem01 |
128 | 4 TiB | 2 x AMD EPYC 7713 | - |
| gpu | gpu01 |
128 | 1 TiB | 2 x Intel Gold 6338 | 4 x NVIDIA A100 80G |
| gpu (x2) | gpu02 & gpu03 |
256 | 1.5 TiB | 2 x AMD EPYC 9554 | 4 x NVIDIA L40S 48G |
| visu | visu01 |
128 | 512 GiB | 2 x AMD EPYC 7513 | 1 x NVIDIA RTX6000 24G |
Disk spaces available¶
| Path | Limit | Writable | Description |
|---|---|---|---|
/home/<username> |
10 GB | Store your configuration files | |
/work/user/<username> |
1 TB | Working directory | |
/save/user/<username> |
250 GB | Data you want to save (replication) with 30 retention days | |
/usr/local/bioinfo/src/ |
- | Directory gathering all bioinformatic softwares (FAQ) | |
/bank |
- | Biological banks in different format (FAQ) |
If you need more space in /work or in /save you are invited to fill the resources request form.
How to get my disk quota and usage?¶
- On
/work:
- On
/saveand/home:
Default CPU time¶
Academic account quota is 100 000 h/per calendar year
Beyond these 100 000 hours, you will need to submit a science project (by the resources request form) to estimate the real needs of the bioinformatics environment.
According to results from this evaluation, but also their geographical and institutional origin, users can then either continue their treatments or be invited to contribute financially to infrastructure, or be redirected to regional or national mésocentres calculation.
Non-academic account quota is 500 h/per calendar year for testing the infrastructure.
Overtime calculation will be charged (price on request).
In order to know your quota, please use the following command:
Default job resource¶
Without any parameters all jobs are limited to:
- 2GB (memory)
- 1 CPU (thread)
- 2 hours (on
workqonly)
CPU (threads) limitation rules¶
It depends on the status of your Slurm group (contributors, INRAe and/or REGION, others).
| Max threads | workq (group) | workq (user) | unlimitq (All users) | unlimitq (user) |
|---|---|---|---|---|
| Contributors | 5184 | 750 | 780 | 500 |
| INRAe/Region | 4409 | 500 | 780 | 376 |
| Others | 1296 | 250 | 780 | 100 |
In order to known the status and the limits of your account, please use:
In this command, $(whoami) will be automagically replaced by your Genotoul Bioinfo login.
If the group limit is reached, your jobs are queued with the reason “AssocGrpCpuLimit”.
Memory limitation rules¶
It depends on the status of your Slurm group (contributors, INRAe and/or REGION, others).
| Max mem | workq (group) | workq (user) | unlimitq (All users) | unlimitq (user) |
|---|---|---|---|---|
| Contributors | 82 TB | 12 TB | - | 8 TB |
| INRAe/Region | 71 TB | 8 TB | - | 6 TB |
| Others | 21 TB | 4 TB | - | 2 TB |
To known the status and the limits of your account:
In this command, $(whoami) will be automagically replaced by your Genotoul Bioinfo login.
If the group limit is reached, your jobs are queued with the reason “AssocGrpMemLimit”.
Job number limitation rules¶
| Case | Limit |
|---|---|
| Max job per user | 2500 |
| Max job for all users | 10000 |
| Max task array per job | 2501 |
Time limitation rules¶
| Slurm Partition | Default Time | Max Time |
|---|---|---|
| workq | 2 hours | 96 hours (4 days) |
| gpuq | 96 hours (4 days) | 96 hours (4 days) |
| interq | 12 hours | 12 hours |
| unlimitq, wflowq | 90 days (3 months) | 90 days (3 months) |
Useful scripts/alias¶
Useful commands summary¶
The useful_commands command will give you a summary of some useful commands.
Useful account informations¶
saccount_info <username> command will gives you some useful informations of your account like:
- account expiration date and last password change date ( every year)
- your primary Linux group
- your secondary Linux groups if you have some
- status of your Linux primary group in Slurm (contributors, inraregion or others)
- your groups' members
- some Slurm limitations of your account
squeue long format¶
sq_long:squeueverbose with detail for:JOBID NAME USER QOS PARTITION NODES CPUS MIN_MEMORY TIME_LIMIT TIME_LEFT STATE NODELIST REASON. Can take all squeue options. See--helpoption for help.sq_debug:squeueverbose for debug. ShowCOMMANDandWORKDIRfor a job. Can take all squeue options. See--helpoption for help.sq_run: sq_debug for running jobs. Can take all squeue options. See--helpoption for help.sq_pend: sq_debug for pending jobs. Can take all squeue options. See--helpoption for help.
sacct long format¶
sa_debug:sacctverbose. Can take allsacctoptions. See--helpoption for help.