系统度量的语义约定¶
NOTICE Semantic Conventions are moving to a new location.
No changes to this document are allowed.
Status: Experimental
This document describes instruments and attributes for common system level metrics in OpenTelemetry. Consider the general metric semantic conventions when creating instruments not explicitly defined in the specification.
- 系统度量的语义约定
- Metric Instruments
system.cpu.
- Processor metricssystem.memory.
- Memory metricssystem.paging.
- Paging/swap metricssystem.disk.
- Disk controller metricssystem.filesystem.
- Filesystem metricssystem.network.
- Network metricssystem.processes.
- Aggregate system process metricssystem.{os}.
- OS Specific System Metrics
Metric Instruments¶
system.cpu.
- Processor metrics¶
Description: System level processor metrics.
Name | Description | Units | Instrument Type (*) | Value Type | Attribute Key(s) | Attribute Values |
---|---|---|---|---|---|---|
system.cpu.time | s | Counter | Double | state | idle, user, system, interrupt, etc. | |
cpu | CPU number [0..n-1] | |||||
system.cpu.utilization | Difference in system.cpu.time since the last measurement, divided by the elapsed time and number of CPUs | 1 | Gauge | Double | state | idle, user, system, interrupt, etc. |
cpu | CPU number (0..n) |
system.memory.
- Memory metrics¶
Description: System level memory metrics. This does not include paging/swap memory.
Name | Description | Units | Instrument Type (*) | Value Type | Attribute Key | Attribute Values |
---|---|---|---|---|---|---|
system.memory.usage | By | UpDownCounter | Int64 | state | used, free, cached, etc. | |
system.memory.utilization | 1 | Gauge | Double | state | used, free, cached, etc. |
system.paging.
- Paging/swap metrics¶
Description: System level paging/swap memory metrics.
Name | Description | Units | Instrument Type (*) | Value Type | Attribute Key | Attribute Values |
---|---|---|---|---|---|---|
system.paging.usage | Unix swap or windows pagefile usage | By | UpDownCounter | Int64 | state | used, free |
system.paging.utilization | 1 | Gauge | Double | state | used, free | |
system.paging.faults | {fault} | Counter | Int64 | type | major, minor | |
system.paging.operations | {operation} | Counter | Int64 | type | major, minor | |
direction | in, out |
system.disk.
- Disk controller metrics¶
Description: System level disk performance metrics.
Name | Description | Units | Instrument Type (*) | Value Type | Attribute Key | Attribute Values |
---|---|---|---|---|---|---|
system.disk.io | By | Counter | Int64 | device | (identifier) | |
direction | read, write | |||||
system.disk.operations | {operation} | Counter | Int64 | device | (identifier) | |
direction | read, write | |||||
system.disk.io_time[1] | Time disk spent activated | s | Counter | Double | device | (identifier) |
system.disk.operation_time[2] | Sum of the time each operation took to complete | s | Counter | Double | device | (identifier) |
direction | read, write | |||||
system.disk.merged | {operation} | Counter | Int64 | device | (identifier) | |
direction | read, write |
1 The real elapsed time ("wall clock") used in the I/O path (time from operations running in parallel are not counted). Measured as:
- Linux: Field 13 from procfs-diskstats
- Windows: The complement of
"Disk\% Idle Time"
performance counter:
uptime * (100 - "Disk\% Idle Time") / 100
2 Because it is the sum of time each request took, parallel-issued requests each contribute to make the count grow. Measured as:
- Linux: Fields 7 & 11 from procfs-diskstats
- Windows: "Avg. Disk sec/Read" perf counter multiplied by "Disk Reads/sec" perf counter (similar for Writes)
system.filesystem.
- Filesystem metrics¶
Description: System level filesystem metrics.
Name | Description | Units | Instrument Type (*) | Value Type | Attribute Key | Attribute Values |
---|---|---|---|---|---|---|
system.filesystem.usage | By | UpDownCounter | Int64 | device | (identifier) | |
state | used, free, reserved | |||||
type | ext4, tmpfs, etc. | |||||
mode | rw, ro, etc. | |||||
mountpoint | (path) | |||||
system.filesystem.utilization | 1 | Gauge | Double | device | (identifier) | |
state | used, free, reserved | |||||
type | ext4, tmpfs, etc. | |||||
mode | rw, ro, etc. | |||||
mountpoint | (path) |
system.network.
- Network metrics¶
Description: System level network metrics.
Name | Description | Units | Instrument Type (*) | Value Type | Attribute Key | Attribute Values |
---|---|---|---|---|---|---|
system.network.dropped[1] | Count of packets that are dropped or discarded even though there was no error | {packet} | Counter | Int64 | device | (identifier) |
direction | transmit, receive | |||||
system.network.packets | {packet} | Counter | Int64 | device | (identifier) | |
direction | transmit, receive | |||||
system.network.errors[2] | Count of network errors detected | {error} | Counter | Int64 | device | (identifier) |
direction | transmit, receive | |||||
system.network.io | By | Counter | Int64 | device | (identifier) | |
direction | transmit, receive | |||||
system.network.connections | {connection} | UpDownCounter | Int64 | device | (identifier) | |
protocol | tcp, udp, etc. | |||||
state | If specified, SHOULD be one of: close, close_wait, closing, delete, established, fin_wait_1, fin_wait_2, last_ack, listen, syn_recv, syn_sent, time_wait. A stateless protocol MUST NOT set this attribute. |
1 Measured as:
- Linux: the
drop
column in/proc/dev/net
(source). - Windows:
InDiscards
/OutDiscards
fromGetIfEntry2
.
2 Measured as:
- Linux: the
errs
column in/proc/dev/net
(source). - Windows:
InErrors
/OutErrors
fromGetIfEntry2
.
system.processes.
- Aggregate system process metrics¶
Description: System level aggregate process metrics. For metrics at the individual process level, see process metrics.
Name | Description | Units | Instrument Type (*) | Value Type | Attribute Key | Attribute Values |
---|---|---|---|---|---|---|
system.processes.count | Total number of processes in each state | {process} | UpDownCounter | Int64 | status | running, sleeping, etc. |
system.processes.created | Total number of processes created over uptime of the host | {process} | Counter | Int64 | - | - |
system.{os}.
- OS Specific System Metrics¶
Instrument names for system level metrics that have different and conflicting
meaning across multiple OSes should be prefixed with system.{os}.
and follow
the hierarchies listed above for different entities like CPU, memory, and
network.
For example, UNIX load average over a given interval is not well standardized and its value across different UNIX like OSes may vary despite being under similar load:
Without getting into the vagaries of every Unix-like operating system in existence, the load average more or less represents the average number of processes that are in the running (using the CPU) or runnable (waiting for the CPU) states. One notable exception exists: Linux includes processes in uninterruptible sleep states, typically waiting for some I/O activity to complete. This can markedly increase the load average on Linux systems.
(source of quote, linux source code)
An instrument for load average over 1 minute on Linux could be named
system.linux.cpu.load_1m
, reusing the cpu
name proposed above and having an
{os}
prefix to split this metric across OSes.