New HTC cluster nodes

Dear HTC cluster users,

We have added 17 nodes to HTC cluster. We have created a separate partition sandy_bridge for the 17 nodes. Each of the 17 nodes has 16 cores, 8 GB/core memory. The original 24 HTC nodes have 16 GB/core memory. The DefMemPerCPU has been changed to 8GB. Use the following slurm parameter to request the proper CPUs and memory for the jobs.

#SBATCH --cpus-per-task=4 # Request that ncpus be allocated per process.
#SBATCH --mem=10g # Memory pool for all cores (see also --mem-per-cpu)

To use this partition sandy_bridge, add the following line to your job submission script.

#SBATCH -p sandy_bridge # partition (queue)

We cannot mount pghbio on these 17 nodes at this stage. These nodes cannot access NCBI now. We planned to solve the two problems by next Tuesday, and we will then merge the sandy_bridge queue to htc queue.

CRC Team