Memory Statistics
Memory Statistics
Plugin: proc.plugin Module: /proc/vmstat
Overview
Linux Virtual memory subsystem.
Information about memory management, indicating how effectively the kernel allocates and frees memory resources in response to system demands.
Monitors page faults, which occur when a process requests a portion of its memory that isn’t immediately available. Monitoring these events can help diagnose inefficiencies in memory management and provide insights into application behavior.
Tracks swapping activity — a vital aspect of memory management where the kernel moves data from RAM to swap space, and vice versa, based on memory demand and usage. It also monitors the utilization of zswap, a compressed cache for swap pages, and provides insights into its usage and performance implications.
In the context of virtualized environments, it tracks the ballooning mechanism which is used to balance memory resources between host and guest systems.
For systems using NUMA architecture, it provides insights into the local and remote memory accesses, which can impact the performance based on the memory access times.
The collector also watches for ‘Out of Memory’ kills, a drastic measure taken by the system when it runs out of memory resources.
This collector is only supported on the following platforms:
- linux
This collector only supports collecting metrics from a single instance of this integration.
Default Behavior
Auto-Detection
This integration doesn’t support auto-detection.
Limits
The default configuration for this integration does not impose any limits on data collection.
Performance Impact
The default configuration for this integration is not expected to impose a significant performance impact on the system.
Setup
Prerequisites
No action required.
Configuration
File
There is no configuration file.
Options
There are no configuration options.
Examples
There are no configuration examples.
Metrics
Metrics grouped by scope.
The scope defines the instance that the metric belongs to. An instance is uniquely identified by a set of labels.
Per Memory Statistics instance
This scope has no labels.
Metrics:
| Metric | Dimensions | Unit |
|---|---|---|
| mem.swapio | in, out | KiB/s |
| system.pgpgio | in, out | KiB/s |
| system.pgfaults | minor, major | faults/s |
| mem.balloon | inflate, deflate, migrate | KiB/s |
| mem.zswapio | in, out | KiB/s |
| mem.ksm_cow | swapin, write | KiB/s |
| mem.thp_faults | alloc, fallback, fallback_charge | events/s |
| mem.thp_file | alloc, fallback, mapped, fallback_charge | events/s |
| mem.thp_zero | alloc, failed | events/s |
| mem.thp_collapse | alloc, failed | events/s |
| mem.thp_split | split, failed, split_pmd, split_deferred | events/s |
| mem.thp_swapout | swapout, fallback | events/s |
| mem.thp_compact | success, fail, stall | events/s |
| mem.oom_kill | kills | kills/s |
| mem.numa | local, foreign, interleave, other, pte_updates, huge_pte_updates, hint_faults, hint_faults_local, pages_migrated | events/s |
Alerts
The following alerts are available:
| Alert name | On metric | Description |
|---|---|---|
| 30min_ram_swapped_out | mem.swapio | percentage of the system RAM swapped in the last 30 minutes |
| oom_kill | mem.oom_kill | number of out of memory kills in the last 30 minutes |