HPC Data Storage Administrator (Research & HPC Data Platforms)

Key Responsibilities

• Architecture: Adapt and evolve the technology designs of existing systems to meet the needs of future computing platforms and research aims.
• Platform Management: Deliver on the scaling, reliability, security, compliance, operations, and lifecycle management of our primary research computing storage platforms, including for high-risk data.
• Tiered Storage Architecture: Oversee the integration of Lustre HSM on the Elm platform, managing data movement policies between parallel filesystems and MinIO object storage.
• Performance Engineering: Tune I/O for large-scale High Performance Computing and AI workloads.
• Community Stewardship: Represent Stanford within the Lustre community and other key community groups, contributing to the upstream roadmap and maintaining a vendor-neutral storage strategy.

Required Qualifications

• Education: Bachelor’s degree and eight years of relevant experience, or a combination of education and relevant experience.
• Expertise at Scale: 8+ years of hands-on experience architecting, building, and managing Lustre and ZFS or similar filesystems at the 20PB+ scale.
• Object Storage & HSM: Deep technical fluency in MinIO and Lustre HSM (copytools, policy engines like RobinHood) or similar tools.
• Kernel & Network Mastery: Expert-level knowledge of the Linux kernel and large-scale InfiniBand/Ethernet fabric tuning.
• In-depth Troubleshooting Experience: Must be capable of leading the debugging of issues such as kernel panics, LNet congestion, and metadata bottlenecks.
• Leadership: Proven experience mentoring junior admins and leading large-scale migration projects without data loss.
• Communication: Strong written and verbal communication skills.

Senior Storage Systems Administrator Key Responsibilities

• Platform Management: Contribute to the scaling, reliability, security, compliance, operations, and lifecycle management of our primary research computing storage platforms, including for high-risk data.
• Operational Excellence: Perform complex filesystem upgrades, kernel patches, and hardware refreshes with minimal downtime.
• Maintenance: Manage the physical and logical health of the storage fleet, including RMA processes, firmware updates, and disk replacement cycles.
• Monitoring & Telemetry: In collaboration with others, build and maintain sophisticated observability stacks for real-time I/O tracking and trend analysis.
• User Support: Act as an escalation point for researchers struggling with complex I/O patterns, job failures, or data access issues.

Qualifications

• Experience: 5+ years of Linux Systems Administration, with 3+ years specifically in an HPC or large-scale data environment.
• Technical Stack: Strong hands-on experience with Lustre, ZFS, MinIO, and/or similar technologies.
• Scripting: Advanced proficiency in scripting languages for automating routine storage tasks and parsing system logs.
• Hardware Mastery: Comfortable with the physical aspects of the role—diagnosing hardware failures and understanding power/cooling requirements for high-density storage.
• Communication: Strong written and verbal communication skills.