...
Ares is built with Infiniband EDR interconnect and nodes of the following specification:
Partition | Number of nodes | CPU | RAM | Accelerator |
---|---|---|---|---|
plgrid and plgrid-* | 532 | 48 cores, Intel(R) Xeon(R) Platinum 8268 CPU @ 2.90GHz | 192GB | |
plgrid-bigmem | 256 | 48 cores, Intel(R) Xeon(R) Platinum 8268 CPU @ 2.90GHz | 384GB | |
plgrid-gpu-v100 | 9 | 32 cores, Intel(R) Xeon(R) Gold 6242 CPU @ 2.80GHz | 384GB | Tesla V100-SXM2 |
Job submission
Jobs Ares is using Slurm resource manager, jobs should be submitted to the following partitions:
Name | Timelimit | Remarks |
---|---|---|
plgrid | 72h | Standard partition. |
plgrid-long | 168h | Used for jobs with extended runtime. |
plgrid-testing | 1h | High priority, testing jobs, limited to 3 jobs. |
plgrid-bigmem | 72h | Jobs using an extended amount of memory. |
plgrid-now | 12h | The highest priority, interactive jobs, limited to 1 running job. |
plgrid-gpu-v100 | 72h | GPU partition. |
Storage
Available storage spaces are described in the following table:
Location | Physical location | Purpose |
---|---|---|
$HOME | /net/people/<login> | Storing own applications, configuration files |
$SCRATCH | /net/pr2/scratch/people/<login> | High-speed storage used for short-lived data heavily used in computations. |
group storage | /net/pr2/projects/plgrid/<group name> | Long term storage, used for data living for the period of computing grant. |
Current usage, capacity and other storage attributes can be checked by issuing the hpc-fs
command.
...
and the environment can be purged by:
module purge
More information
Ares is following Prometheus' configuration and usage patterns. Prometheus documentation can be found here: https://kdm.cyfronet.pl/portal/Prometheus:Basics