📁 memory.stat

This file provides detailed statistics about memory usage in the cgroup. It includes various counters such as:

  • cache: Amount of memory used for caching files.
  • rss: Resident Set Sise – non-swapped physical memory.
  • mapped_file: Memory mapped files.
  • pgfault / pgmajfault: Page faults and major page faults.
  • inactive_anon, active_anon: Anonymous memory pages.
  • hierarchical_memory_limit: Effective memory limit including parent cgroups.
  • total_rss, total_cache: Aggregated values including child cgroups.

Use case: Helps identify what kind of memory is being used and whether the workload is memory-intensive or leaking memory.

📁 memory.failcnt

This file shows how many times the memory usage in the cgroup has exceeded the limit set by memory.limit_in_bytes.

  • A non-zero value indicates that the process tried to use more memory than allowed and was denied.
  • This is a strong indicator of memory starvation.

Use case: If NRPE or sssd has a high failcnt, it means they are being throttled or killed due to memory limits.

📁 memory.usage_in_bytes

This file shows the current memory usage by the cgroup in bytes.

  • It includes both anonymous and file-backed memory.
  • Compare this with memory.limit_in_bytes to see how close the process is to its cap.

Use case: Helps you monitor real-time memory consumption and adjust limits accordingly.

🔧 Example: Checking NRPE Memory Usage

CGROUP_PATH="/sys/fs/cgroup/memory/system.slice/nrpe.service"
cat $CGROUP_PATH/memory.usage_in_bytes
cat $CGROUP_PATH/memory.failcnt
cat $CGROUP_PATH/memory.stat

Updated: