Node exporter metric definition file (metrics_node_exporter.conf)
- Organization of this page
Format
[ { "name":"trend-data-metric-names", "default":default-selection-state, "promql":"PromQL-statement", "resource_en":{ "category":"metric-category-(English)", "label":"metric-display-name-(English)", "description":"metric-description-(English)", "unit":"metric-unit-(English)" }, "resource_ja":{ "category":"metric-category-(Japanese)", "label":"metric-display-name-(Japanese)", "description":"metric-description-(Japanese)", "unit":"metric-unit-(Japanese)" }, "drop_legend_labels": ["label-name", ...] }, ... ]#
- #
-
The number of elements that can be described in [] is 1 to 1,000. If an out-of-range number of elements is described, a KAJY24609-E error message is output.
File
metrics_node_exporter.conf
metrics_node_exporter.conf.model (model file)
Storage directory
■Integrated manager host
- In Windows:
-
-
For a physical host
Manager-path\conf\imdd\plugin\jp1pccs\
-
For a logical host
shared-folder\jp1imm\conf\imdd\plugin\jp1pccs\
-
- In Linux:
-
-
For a physical host
/etc/opt/jp1imm/conf/imdd/plugin/jp1pccs/
-
For a logical host
shared-directory/jp1imm/conf/imdd/plugin/jp1pccs/
-
Description
This file defines the metric information for the Node exporter to be displayed on the Trend tab of the Unified Operations Viewer screen.
The defined contents are used for the return values of __metricListGet method and __timeSeriesDataGet method of the JP1/IM - Agent product plug-in.
If JP1/IM - Manager is in a hierarchical configuration and you want to refer to trend data stored in Trend data Management Database of lower manager from Integrated manager, you must add trend data metrics that you want to refer to metric definition file of Integrated manager.
Character code
UTF-8 (without BOM)
Line feed code
In Windows: CR+LF
In Linux: LF
When the definitions are applied
It is reflected when metric information is obtained on the Trend tab of the Integrated Operation Viewer screen or with the REST API.
Information that is specified
Member name |
Optional |
Type |
Value to set |
|
---|---|---|---|---|
name |
No |
string |
Set trend data metric names. The metric name of trend data (time series data) is displayed as a character string indicating the type of trend data on the Trend tab of the Unified Operation Viewer screen. 1 to 255 characters, and the following characters can be specified.
If it contains characters that cannot be specified, a KAJY24604-E or KAJY24605-E error message is printed. The metric name must be unique within the metric definition file. Otherwise, a KAJY24608-E error message is output. |
|
default |
Yes |
boolean |
Set default selection state. Specifies whether the check box of trend data selected on the Trend tab of the Integrated Operation Viewer screen is selected by default.
If it is omitted, it is not checked by default. |
|
promql |
No |
string |
Set PromQL statement. You can configure PromQL statements that can be used in JP1/IM trend data reference API. The string "$jp1im_TrendData_labels" in the PromQL statement is replaced by the PromQL statement described in "■Replacing $jp1im_TrendData_labels" below. For notes on PromQL statements, see Note on PromQL expression. Specify 1 to 1,023 characters except control characters (0x00~0x1F, 0x7F~0x9F). If it contains characters that cannot be specified, a KAJY24604-E or KAJY24605-E error message is printed. |
|
resource_en |
No |
object |
Configure resource information for metrics. |
|
category |
Yes |
string |
Set the category of the metric. Specify 1 to 255 characters except control characters (0x00 to 0x1F, 0x7F to 0x9F). If it contains characters that cannot be specified, a KAJY24604-E or KAJY24605-E error message is printed. The default is not to set the metric category. |
|
label |
Yes |
string |
Set the display name of the metric. Specify 1 to 255 characters except control characters (0x00 to 0x1F, 0x7F to 0x9F). If it contains characters that cannot be specified, a KAJY24604-E or KAJY24605-E error message is printed. By default, the settings of the name member are used for the display name of the metric on the Trends tab of the Unified Operations Viewer screen. Also, the display name of the metric is not returned by the metric list acquisition API. |
|
description |
Yes |
string |
Sets the description of the metric. Specifies 1 to 1,023 characters, excluding control characters (0x00 to 0x1F,0x7F to 0x9F). If it contains characters that cannot be specified, a KAJY24604-E or KAJY24605-E error message is printed. By default, metric descriptions are not displayed on the Trends tab of the Unified Operations Viewer screen. Also, the metric description is not returned by the metric list acquisition API. |
|
unit |
No |
string |
Set the units for the metric. Specify 1 to 255 characters except control characters (0x00 to 0x1F, 0x7F to 0x9F). If it contains characters that cannot be specified, a KAJY24604-E or KAJY24605-E error message is printed. |
|
resource_ja |
Yes |
object |
Set the resource information (Japanese) for the metric. By default, the resource_en settings are used. |
|
category |
Yes |
string |
Set the category of the metric (Japanese). Specify 1 to 255 characters except control characters (0x00 to 0x1F, 0x7F to 0x9F). If it contains characters that cannot be specified, a KAJY24604-E or KAJY24605-E error message is printed. By default, the settings of the resource_en category are used. |
|
label |
Yes |
string |
Set the display name (Japanese) of the metric. Specify 1 to 255 characters except control characters (0x00 to 0x1F, 0x7F to 0x9F). If it contains characters that cannot be specified, a KAJY24604-E or KAJY24605-E error message is printed. If it is the default, the label setting of the resource_en is used. |
|
description |
Yes |
string |
Set the description of the metric (Japanese). Specifies 1 to 1,023 characters, excluding control characters (0x00 to 0x1F,0x7F to 0x9F). If it contains characters that cannot be specified, a KAJY24604-E or KAJY24605-E error message is printed. By default, the description setting of the resource_en is used. |
|
unit |
Yes |
string |
Set the unit of the metric (Japanese). Specify 1 to 255 characters except control characters (0x00 to 0x1F, 0x7F to 0x9F). If it contains characters that cannot be specified, a KAJY24604-E or KAJY24605-E error message is printed. By default, the settings of the resource_en unit are used. |
|
drop_legend_labels |
Yes |
Array (string) |
Set the label name if you want to add a label that you want to exclude from the trend chart legend. In addition to the labels to be excluded shown in 3.15.6(4)(b) Instance name (a string to display as a legend in the graph) of the JP1/Integrated Management 3 - Manager Overview and System Design Guide, the names of labels to be excluded from the legend of the trend graph are described. Specify 1 to 255 characters except control characters (0x00 to 0x1F, 0x7F to 0x9F). If it contains characters that cannot be specified, it is not excluded from the legend of the trend graph. You can specify up to 100 labels. If there are more than 100 labels, the 101st and subsequent labels are ignored. |
- ■Replacing $jp1im_TrendData_labels
-
The string "$jp1im_TrendData_labels" in the promql value is replaced with a PromQL statement to narrow down the target of retrieval when retrieving performance data.
The following is the replaced PromQL statement for each SID type of configuration information specified in the Trending API or Unified Operations Viewer (for CloudWatchSIDs other than EC2, for each monitored AWS namespace). The bold part (variable value) is replaced by the Exporter setting managed by the corresponding SID or the sample value output by the Exporter.
Table 2‒95: PromQL statement after replacing $jp1im_TrendData_labels Configuration information SID type
PromQL statement after replacement
Agent SID
{jp1_pc_prome_hostname="Prometheus-host-name",job="scrape-job-name",instance="value-of-instance-label"}
Remote agent SID
CloudWatchSID of EC2
{jp1_pc_prome_hostname="Prometheus-host-name",job="scrape-job-name",instance="value-of-instance-label"},jp1_pc_nodelabel="The value that you set for the jp1_pc_nodelabel tag of the resource in AWS"}
CloudWatchSID other than EC2
Model file settings (initial state)
The setting contents (initial state) of each metric described in the model file of the Node exporter metric definition file are shown below.
-
cpu_used_rate#
■Configuration contents (default status)
Member name
Configuration contents (default status)
name
cpu_used_rate
default
true
The promql for metric definition File (including $jp1im_TrendData_labels)
(avg by (instance,job,jp1_pc_nodelabel,jp1_pc_prome_hostname) (rate(node_cpu_seconds_total{mode=\"system\"}[2m]) and $jp1im_TrendData_labels) + avg by (instance,job,jp1_pc_nodelabel,jp1_pc_prome_hostname) (rate(node_cpu_seconds_total{mode=\"user\"}[2m]) and $jp1im_TrendData_labels)) * 100
resource_en
category
platform_unix
label
CPU used rate
description
CPU usage.It also indicates the average value per processor. [Units: %]
unit
%
resource_ja
category
platform_unix
label
CPU使用率
description
CPU使用率(%)。
プロセッサごとの割合の平均値でもある。
unit
%
- #
-
Equivalent to the CPU % field in the PI record of JP1/PFM - Agent for Platform (Unix).
-
memory_unused#
■Configuration contents (default status)
Member name
Configuration contents (default status)
name
memory_unused
default
false
The promql for metric definition File (including $jp1im_TrendData_labels)
(node_memory_MemAvailable_bytes and $jp1im_TrendData_labels)/1024/1024/1024
resource_en
category
platform_unix
label
Memory unused
description
Size of the physical memory that can actually be used by the application.
unit
GB
resource_ja
category
platform_unix
label
空きメモリ量
description
実際にアプリケーションが使用することができる物理メモリーのサイズ。
unit
ギガバイト
- #
-
Equivalent to the Effective Free Mem Mbytes field in the PI record of JP1/PFM - Agent for Platform (Unix).
However, the number unit are different.
-
memory_unused_rate#
■Configuration contents (default status)
Member name
Configuration contents (default status)
name
memory_unused_rate
default
false
The promql for metric definition File (including $jp1im_TrendData_labels)
(node_memory_MemAvailable_bytes and $jp1im_TrendData_labels) / (node_memory_MemTotal_bytes and $jp1im_TrendData_labels) * 100
resource_en
category
platform_unix
label
Available memory percentage
description
Percentage of physical memory actually available to the application
unit
%
resource_ja
category
platform_unix
label
空きメモリ率
description
実際にアプリケーションが使用することができる物理メモリーの割合。
unit
%
- #
-
Equivalent to Effective Free Mem % field in PI record of JP1/PFM - Agent for Platform (Unix).
-
disk_unused#
■Configuration contents (default status)
Member name
Configuration contents (default status)
name
disk_unused
default
false
The promql for metric definition File (including $jp1im_TrendData_labels)
(node_filesystem_free_bytes and $jp1im_TrendData_labels)/(1024*1024*1024)
resource_en
category
platform_unix
label
Disk unused
description
Size of the unused area of the disk. [Units: GB]
unit
GB
resource_ja
category
platform_unix
label
空きディスク領域
description
ディスクの未使用領域のサイズ。(単位:ギガバイト)
unit
ギガバイト
- #
-
Equivalent to the Mbytes Free field in the PD_FSL record of JP1/PFM - Agent for Platform (Unix).
However, the number unit are different.
-
disk_unused_rate#
■Configuration contents (default status)
Member name
Configuration contents (default status)
name
disk_unused_rate
default
false
The promql for metric definition File (including $jp1im_TrendData_labels)
(node_filesystem_free_bytes and $jp1im_TrendData_labels) / (node_filesystem_size_bytes and $jp1im_TrendData_labels) * 100
resource_en
category
platform_unix
label
Unused disk space percentage
description
Percentage of unused disk space
unit
%
resource_ja
category
platform_unix
label
空きディスク率
description
ディスクの未使用領域の割合。
unit
%
- #
-
Equivalent to Mbytes Free % field of the PD_FSL record in JP1/PFM - Agent for Platform (Unix).
-
disk_busy_rate#
■Configuration contents (default status)
Member name
Configuration contents (default status)
name
disk_busy_rate
default
false
The promql for metric definition File (including $jp1im_TrendData_labels)
(rate(node_disk_io_time_seconds_total[2m]) and $jp1im_TrendData_labels)*100
resource_en
category
platform_unix
label
Disk busy rate
description
Percentage of time the disk was busy with read and write requests. This value may exceed 100 when processes are continuously executed on a device. [Units: %]
unit
%
resource_ja
category
platform_unix
label
ディスクビジー率
description
ディスクのビジー率(%)。
デバイスに対する処理が連続で行われる場合に「100」を超えることがあります。
unit
%
- #
-
Equivalent to the Busy % field in the PI_DEVD record of JP1/PFM - Agent for Platform (Unix).
-
disk_read_latency
■Configuration contents (default status)
Member name
Configuration contents (default status)
name
disk_read_latency
default
false
The promql for metric definition File (including $jp1im_TrendData_labels)
(rate(node_disk_read_time_seconds_total[2m]) and $jp1im_TrendData_labels) / (rate(node_disk_reads_completed_total[2m]) and $jp1im_TrendData_labels)
resource_en
category
platform_unix
label
Disk read latency
description
Average time (in seconds) taken to perform a single disk read operation
unit
second
resource_ja
category
platform_unix
label
ディスク読み込みレイテンシー
description
1回あたりのディスク読み込みにかかった平均時間(秒)。
unit
秒
-
disk_write_latency
■Configuration contents (default status)
Member name
Configuration contents (default status)
name
disk_write_latency
default
false
The promql for metric definition File (including $jp1im_TrendData_labels)
(rate(node_disk_write_time_seconds_total[2m]) and $jp1im_TrendData_labels) / (rate(node_disk_writes_completed_total[2m]) and $jp1im_TrendData_labels)
resource_en
category
platform_unix
label
Disk write latency
description
Average time (in seconds) taken to perform a single disk write operation
unit
second
resource_ja
category
platform_unix
label
ディスク書き込みレイテンシー
description
1回あたりのディスク書き込みにかかった平均時間(秒)。
unit
秒
-
disk_io_latency
■Configuration contents (default status)
Member name
Configuration contents (default status)
name
disk_io_latency
default
false
The promql for metric definition File (including $jp1im_TrendData_labels)
((rate(node_disk_read_time_seconds_total[2m]) and $jp1im_TrendData_labels) + (rate(node_disk_write_time_seconds_total[2m]) and $jp1im_TrendData_labels)) / ((rate(node_disk_reads_completed_total[2m]) and $jp1im_TrendData_labels) + (rate(node_disk_writes_completed_total[2m]) and $jp1im_TrendData_labels))
resource_en
category
platform_unix
label
Disk I/O latency
description
Average time (in seconds) taken to perform a single disk write and read operation
unit
second
resource_ja
category
platform_unix
label
ディスクIOレイテンシー
description
1回あたりのディスク書き込みとディスク読み込みにかかった平均時間(秒)。
unit
秒
-
network_sent#
■Configuration contents (default status)
Member name
Configuration contents (default status)
name
network_sent
default
false
The promql for metric definition File (including $jp1im_TrendData_labels)
(rate(node_network_transmit_packets_total[2m]) and $jp1im_TrendData_labels)
resource_en
category
platform_unix
label
Network sent
description
Rate at which packets were sent throughthe network interface. [Units: packets/second]
unit
packets/second
resource_ja
category
platform_unix
label
ネットワークの送信速度
description
ネットワークインターフェースで送信されるパケットの割合。(単位:パケット/秒)
unit
パケット/秒
- #
-
Equivalent to the Pkts Xmitd/sec field in the PI_NIND record of JP1/PFM - Agent for Platform (Unix).
-
network_received#
■Configuration contents (default status)
Member name
Configuration contents (default status)
name
network_received
default
false
The promql for metric definition File (including $jp1im_TrendData_labels)
(rate(node_network_receive_packets_total[2m]) and $jp1im_TrendData_labels)
resource_en
category
platform_unix
label
Network received
description
Rate at which packets were receivedthrough the network interface. [Units: packets/second]
unit
packets/second
resource_ja
category
platform_unix
label
ネットワークの受信速度
description
ネットワークインターフェースで受信されるパケットの割合。(単位:パケット/秒)
unit
パケット/秒
- #
-
Equivalent to the Pkts Rcvd/sec field in the PI_NIND record of JP1/PFM - Agent for Platform (Unix).