Pseudonymization at Scale: OLCF's Summit Usage Data Case Study

TL;DR

This paper presents a scalable pseudonymization workflow for large HPC system log datasets, enabling privacy-preserving data sharing while maintaining data utility, demonstrated on the OLCF Summit supercomputer's user data.

Contribution

The paper introduces a parallelized pseudonymization workflow for large-scale HPC log data, improving processing efficiency and enabling open data sharing with privacy considerations.

Findings

Workflow reduces processing time from 20+ hours to around 2 hours.

Demonstrates scalability of pseudonymization on multiple HPC systems.

Publishes the pseudonymized dataset and workflow for community use.

Abstract

The analysis of vast amounts of data and the processing of complex computational jobs have traditionally relied upon high performance computing (HPC) systems. Understanding these analyses' needs is paramount for designing solutions that can lead to better science, and similarly, understanding the characteristics of the user behavior on those systems is important for improving user experiences on HPC systems. A common approach to gathering data about user behavior is to analyze system log data available only to system administrators. Recently at Oak Ridge Leadership Computing Facility (OLCF), however, we unveiled user behavior about the Summit supercomputer by collecting data from a user's point of view with ordinary Unix commands. Here, we discuss the process, challenges, and lessons learned while preparing this dataset for publication and submission to an open data challenge. The original dataset contains personal identifiable information (PII) about OLCF users which needed be masked prior to publication, and we determined that anonymization, which scrubs PII completely, destroyed too much of the structure of the data to be interesting for the data challenge. We instead chose to pseudonymize the dataset to reduce its linkability to users' identities. Pseudonymization is significantly more computationally expensive than anonymization, and the size of our dataset, approximately 175 million lines of raw text, necessitated the development of a parallelized workflow that could be reused on different HPC machines. We demonstrate the scaling behavior of the workflow on two leadership class HPC systems at OLCF, and we show that we were able to bring the overall makespan time from an impractical 20+ hours on a single node down to around 2 hours. As a result of this work, we release the entire pseudonymized dataset and make the workflows and source code publicly available.