USF HII Compute Portal

HII-HPC Cluster

HII-HPC Cluster

Frequently Asked Questions

1. What is the purpose of the HII-HPC Cluster?

The HII-HPC Cluster is a specialized set of compute resources offered to facilitate Bioinformatics research for large-scale workloads.

2. What is the relationship between the HII-HPC Cluster and CIRCE?

CIRCE, the “Central Instructional and Research Computing Environment”, is the high-performance computing infrastructure provided by USF Research Computing for all faculty, students, staff, and partners.

The HII-HPC Cluster represents a specialized subset of resources integrating into the overall CIRCE environment providing the unique requirements of high-throughput bioinformatics as a workload.

Due to this specialization, please use the HII Compute Portal Documentation as the authoritative source of information for the HII-HPC Cluster.

3. What resources are available on the HII-HPC Cluster?

Since resources are being added to the cluster on a continual basis, the most accurate way to determine available resources is by logging into a head node, hii.rc.usf.edu or hii2.rc.usf.edu, and referring to the sinfo example Show All HII-HPC Resources to obtain the most current information.

4. Can software package “X” be installed on the HII-HPC Cluster?

There are a number of USF RC pre-built packages available on the HII-HPC cluster via Modules.

For Python packages, use a Virtual Environment so you can control your own Python packages.

For R, most packages can be installed into your local R environment following similar patterns to those detailed at https://www.osc.edu/documentation/howto/install-local-R-packages.

For other software:

  1. Compile it yourself using gcc or an Intel Compiler icc (see Modules) configuring and deploying to an installation path under your $HOME or preferably your group share (e.g. /home/d/dvader/software/<name>-<version>/... or /shares/hii-jedi/software/<name>-<version>/...)

  2. Request the software via the USF Help Desk to be added to the Modules environment.

5. Can we install or access a database or other persistent service on the HII-HPC Cluster?

Persistent services are long-running software applications intended to persist state beyond the life of a computational pipeline submitted to the HII-HPC cluster (e.g. mysql, postgresql, etc.).

These services are not currently supported on the HII-HPC cluster however network connections to external services from the head and compute nodes are possible.

Please add the following IP addresses to the external service’s firewall ruleset to allow incoming access if necessary:

  • 131.247.250.90 (Public IP for head node hii.rc.usf.edu)
  • 131.247.250.99 (Public IP for head node hii2.rc.usf.edu)
  • 131.247.244.250 (Public IP for all compute nodes)