DAMON Memory Management Subsystem Adds Tiering, THP Monitoring in Major 2026 Update

By

SeongJae Park, the creator of the Linux kernel's DAMON subsystem, announced a sweeping set of new capabilities at the 2026 Linux Storage, Filesystem, Memory Management, and BPF Summit, solidifying the tool's role as a critical component for system memory oversight.

DAMON, which provides user-space monitoring and management of system memory, is gaining support for memory tiering, data attributes monitoring, and transparent huge pages (THP) handling, among other features.

“These additions represent a fundamental shift in how we can observe and control memory behavior at scale,” Park said during his presentation. “Tiering alone opens up new optimization paths for heterogeneous memory systems.”

Background

DAMON was introduced to the kernel several years ago as a lightweight framework for tracking memory access patterns. It allows administrators and applications to make intelligent decisions about memory allocation, page migration, and reclaim.

DAMON Memory Management Subsystem Adds Tiering, THP Monitoring in Major 2026 Update

Over successive releases, the subsystem has evolved from a basic monitoring tool into a full-fledged management engine, capable of proactive adjustments to memory pressure and workload demands.

What’s New in 2026

The update includes explicit support for memory tiering, enabling DAMON to identify hot and cold pages across different memory types such as DRAM, persistent memory, and CXL-attached memory. This allows the kernel to migrate data to the most cost-effective tier without user intervention.

Data attributes monitoring, another headline feature, lets DAMON track custom metrics defined by the user, extending beyond simple access frequencies to include factors like reuse distance or latency sensitivity.

Transparent huge pages (THP) are also now under DAMON’s purview. “We’re adding a THP monitoring mode that reports fragmentation and collapse opportunities,” Park explained. “It helps system administrators decide when to defragment.”

What This Means

For data center operators and cloud providers, these capabilities translate to more efficient memory utilization, reduced total cost of ownership, and improved application performance predictability.

Memory tiering, in particular, is expected to become a cornerstone of future Linux deployments as hardware diversity increases. The ability to monitor THP health natively also reduces the need for external profiling tools.

“DAMON is no longer just a debugger — it’s a runtime memory optimizer,” said kernel developer Mel Gorman, who attended the session. “This update makes it production-ready for large-scale systems.”

Park noted that all features are already in the development branch and are expected to land in the mainline kernel for the 2026 release cycle.

Developers are encouraged to test the new interfaces via the DAMON user-space library and provide feedback on the kernel mailing list.

The full presentation slides are available from the summit’s online archive.

Tags:

Related Articles

Recommended

Discover More

How to Leverage Your IDE as an AI Quality Variable: A Step-by-Step Guide10 Critical npm Security Risks and How to Mitigate Them (Updated 2025)Decoding the Learning Dynamics of Word2vec: A Mathematical Perspective10 Critical Insights into the RAM and Storage Shortage CrisisHow to Fortify Your Canvas Login Portals Against Mass Extortion Attacks: A Guide Inspired by the ShinyHunters Incident