Why Linux ‘Ate My RAM’: Understanding free, buffers, and cache

When Windows Explorer shows memory usage above 80%, running large programs feels sluggish because swapping occurs; on Linux, the free command often shows high memory usage even when few applications run, which is due to Linux’s memory management design.

free command introduction

Running free -m on a lab file server yields output such as:

Mem: total=3920MB, used=1938MB, free=1982MB, shared=0MB, buffers=497MB, cached=1235MB

Buffers are data waiting to be written to disk, while cache holds data read from disk for future use; both improve I/O performance.

Subtracting buffers and cache from used memory gives the amount actually occupied by applications (205 MB), and adding them to free memory gives the true amount of memory available to programs (3714 MB).

Swap space is unused in this example, indicating sufficient RAM.

Memory classification

From the OS and user perspectives, memory can be categorized as:

Used : memory taken by applications.

Free (available) : memory not used for anything.

Buffers/Cache : memory used by the kernel for I/O caching; it can be reclaimed quickly for applications, so users often consider it free.

Therefore, the third line of free output (‑/‑ buffers/cache) shows memory usage from the user’s viewpoint.

Benefits of buffers/cached

Linux employs buffer cache and page cache to speed up disk I/O. Data read from slow block devices is kept in memory, allowing subsequent accesses to bypass the disk. When memory becomes scarce, the kernel can drop cached pages instead of swapping, preserving performance.

Experiment proof

An experiment reads a 1 GB file twice:

Generate a 1 GB file.

Clear the cache.

Read the file and measure time (≈18 s).

Read the file again (≈0.3 s), a 60‑fold speedup thanks to caching.

References

Discussion on “Linux ate my RAM” (Google Groups).

linuxatemyram.com article.

Additional experiments on cache impact.

Blog post introducing the free command.