About the role
High-Performance Computing (HPC) Administrator
You will be responsible for managing high-density compute nodes, InfiniBand interconnects, job scheduling environments, ensuring high availability, scalability, and efficiency across the cluster ecosystem.
Key Responsibilities
Linux Skills
- Familiarity with one, or more, shell environments (BASH, sh, tcsh, ksh, etc.)
- Experience administering Linux (boot process, file systems, kernel modules)
- Familiarity with core OS services such as: SSH, telnet, FTP, NFS, DNS, DHCP, Samba, LDAP
- Familiarity with networking tools (ping, tracert, tracemon, tcpdump, etc.)
- Understanding of the OSI model and related concepts
Cluster Administration
- Install, configure, and maintain Linux-based compute and login nodes.
- Manage workload schedulers (e.g.,
Slurm
, PBS Pro
) and optimize job throughput and queue performance. - Oversee software provisioning, OS image deployment, and firmware updates across thousands of nodes.
- Manage HPC management frameworks such as
HPE Performance Cluster Manager (HPCM),
Bright Cluster Manager
, or equivalent.
Networking & Interconnect
- Monitor and tune high-speed fabrics such as
InfiniBand (HDR/NDR)
including subnet managers (UFM, OpenSM) and link diagnostics. - Need to ensure low latency, congestion-free data paths across compute and storage.
Monitoring & Automation
- Implement comprehensive cluster monitoring using tools like
Grafana
, Prometheus
, Nagios
or Opsramp
. - Develop automation and orchestration scripts
(Bash, Python, Ansible)
for provisioning, health checks, and performance tuning. - Proactively identify performance bottlenecks and propose corrective actions.
Security & Compliance
- Maintain secure authentication, role-based access, and audit controls across nodes.
- Apply patches, enforce configuration management, and ensure compliance with organizational IT policies.
Qualifications
Essential:
- Bachelors degree in Computer Science, Engineering or a related technical field.
- Strong experience in
Linux systems administration (RHEL/SLES/Ubuntu)
in an HPC or data center environment. - Minimum 5 years of experience in HPC administration is valuable.
- Hands-on expertise with
Slurm
, PBS Pro
job schedulers. - Familiarity with
InfiniBand networking
- Proficiency in scripting (Bash, Python, or Ansible) and version control (Git).
- Good knowledge about the HPE Proliant Gen 10 or Gen11 servers is preferable.
Desirable:
- Experience with
containerized HPC (Singularity/Apptainer/Docker)
and Kubernetes integration
. - Basic understanding of VMware is good.
- Familiarity with Lustre
/GPFS filesystems
, - Certifications like RHCSA
/RHCE
, or CKA or vendor-specific HPC administration credentials.