This white paper discusses the current state of Privileged Access Management (PAM) and its application to High-Performance Computing (HPC) environments. In it, we draw a parallel between HPC and Hadoop since the latter has attracted a lot of attention from PAM vendors, much of which you can apply to HPC either in its more classic on-premises configuration or in evolving hybrid on-premises/cloud configurations.
HPC has been with us for decades. Its roots in supercomputing, where it was available only to a select few with deep pockets, have evolved to capitalize on commodity hardware, making it available to just about any organization cost-effectively.
Over the years, we’ve seen tremendous growth in cloud-based technologies. One such technology is Hadoop, which has attracted many big names such as Facebook, Alibaba, eBay, and Amazon. Whether in public or private clouds, PAM has played a significant role in helping secure Big Data projects that rely on these Hadoop concepts.
This adoption, however, doesn’t mean that classic data center-hosted HPC environments have simply shut down and migrated wholesale to the cloud. Thousands still exist globally in government, industry, and academia. Many of these continue to thrive and grow, serving the needs of disciplines such as bioscience, physics, and climate.
Architecture aside, from a PAM perspective the primary objectives and use cases are identical; protect access to resources and data by preventing the leading cause of breaches — privileged access abuse. Classic HPC can, therefore, benefit significantly from the same PAM advancements designed for Hadoop in the modern era.
This paper discusses these PAM-related risks and how you can leverage modern PAM to address them.