Migrate dashboards to new grafonnet library (#3147)
Some checks failed
golangci-lint / lint (push) Has been cancelled

Migrated away from deprecated Grafonnet library. This replaces panels using Angular JS which are disabled by default in Grafana 11 and will be unsupported in Grafana 12.

Fixes #3046

---------

Signed-off-by: Tom <12222103+critchtionary@users.noreply.github.com>
This commit is contained in:
Tom 2024-12-19 15:49:22 +00:00 committed by GitHub
parent ff97e35a71
commit d0c1d00d18
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
3 changed files with 871 additions and 894 deletions

View file

@ -1,201 +1,178 @@
local grafana = import 'github.com/grafana/grafonnet-lib/grafonnet/grafana.libsonnet';
local grafana = import 'github.com/grafana/grafonnet/gen/grafonnet-latest/main.libsonnet';
local dashboard = grafana.dashboard;
local row = grafana.row;
local prometheus = grafana.prometheus;
local template = grafana.template;
local graphPanel = grafana.graphPanel;
local variable = dashboard.variable;
local row = grafana.panel.row;
local prometheus = grafana.query.prometheus;
local timeSeriesPanel = grafana.panel.timeSeries;
local tsOptions = timeSeriesPanel.options;
local tsStandardOptions = timeSeriesPanel.standardOptions;
local tsQueryOptions = timeSeriesPanel.queryOptions;
local tsCustom = timeSeriesPanel.fieldConfig.defaults.custom;
local tsLegend = tsOptions.legend;
local c = import '../config.libsonnet';
local datasourceTemplate = {
current: {
text: 'default',
value: 'default',
},
hide: 0,
label: 'Data Source',
name: 'datasource',
options: [],
query: 'prometheus',
refresh: 1,
regex: '',
type: 'datasource',
};
local datasource = variable.datasource.new(
'datasource', 'prometheus'
);
local tsCommonPanelOptions =
variable.query.withDatasourceFromVariable(datasource)
+ tsCustom.stacking.withMode('normal')
+ tsCustom.withFillOpacity(100)
+ tsCustom.withShowPoints('never')
+ tsLegend.withShowLegend(false)
+ tsOptions.tooltip.withMode('multi')
+ tsOptions.tooltip.withSort('desc');
local CPUUtilisation =
graphPanel.new(
timeSeriesPanel.new(
'CPU Utilisation',
datasource='$datasource',
span=6,
format='percentunit',
stack=true,
fill=10,
legend_show=false,
) { tooltip+: { sort: 2 } };
)
+ tsCommonPanelOptions
+ tsStandardOptions.withUnit('percentunit');
local CPUSaturation =
// TODO: Is this a useful panel? At least there should be some explanation how load
// average relates to the "CPU saturation" in the title.
graphPanel.new(
timeSeriesPanel.new(
'CPU Saturation (Load1 per CPU)',
datasource='$datasource',
span=6,
format='percentunit',
stack=true,
fill=10,
legend_show=false,
) { tooltip+: { sort: 2 } };
)
+ tsCommonPanelOptions
+ tsStandardOptions.withUnit('percentunit');
local memoryUtilisation =
graphPanel.new(
timeSeriesPanel.new(
'Memory Utilisation',
datasource='$datasource',
span=6,
format='percentunit',
stack=true,
fill=10,
legend_show=false,
) { tooltip+: { sort: 2 } };
)
+ tsCommonPanelOptions
+ tsStandardOptions.withUnit('percentunit');
local memorySaturation =
graphPanel.new(
timeSeriesPanel.new(
'Memory Saturation (Major Page Faults)',
datasource='$datasource',
span=6,
format='rds',
stack=true,
fill=10,
legend_show=false,
) { tooltip+: { sort: 2 } };
)
+ tsCommonPanelOptions
+ tsStandardOptions.withUnit('rds');
local networkOverrides = tsStandardOptions.withOverrides(
[
tsStandardOptions.override.byRegexp.new('/Transmit/')
+ tsStandardOptions.override.byRegexp.withPropertiesFromOptions(
tsCustom.withTransform('negative-Y')
),
]
);
local networkUtilisation =
graphPanel.new(
timeSeriesPanel.new(
'Network Utilisation (Bytes Receive/Transmit)',
datasource='$datasource',
span=6,
format='Bps',
stack=true,
fill=10,
legend_show=false,
)
.addSeriesOverride({ alias: '/Receive/', stack: 'A' })
.addSeriesOverride({ alias: '/Transmit/', stack: 'B', transform: 'negative-Y' })
{ tooltip+: { sort: 2 } };
+ tsCommonPanelOptions
+ tsStandardOptions.withUnit('Bps')
+ networkOverrides;
local networkSaturation =
graphPanel.new(
timeSeriesPanel.new(
'Network Saturation (Drops Receive/Transmit)',
datasource='$datasource',
span=6,
format='Bps',
stack=true,
fill=10,
legend_show=false,
)
.addSeriesOverride({ alias: '/ Receive/', stack: 'A' })
.addSeriesOverride({ alias: '/ Transmit/', stack: 'B', transform: 'negative-Y' })
{ tooltip+: { sort: 2 } };
+ tsCommonPanelOptions
+ tsStandardOptions.withUnit('Bps')
+ networkOverrides;
local diskIOUtilisation =
graphPanel.new(
timeSeriesPanel.new(
'Disk IO Utilisation',
datasource='$datasource',
span=6,
format='percentunit',
stack=true,
fill=10,
legend_show=false,
) { tooltip+: { sort: 2 } };
)
+ tsCommonPanelOptions
+ tsStandardOptions.withUnit('percentunit');
local diskIOSaturation =
graphPanel.new(
timeSeriesPanel.new(
'Disk IO Saturation',
datasource='$datasource',
span=6,
format='percentunit',
stack=true,
fill=10,
legend_show=false,
) { tooltip+: { sort: 2 } };
)
+ tsCommonPanelOptions
+ tsStandardOptions.withUnit('percentunit');
local diskSpaceUtilisation =
graphPanel.new(
timeSeriesPanel.new(
'Disk Space Utilisation',
datasource='$datasource',
span=12,
format='percentunit',
stack=true,
fill=10,
legend_show=false,
) { tooltip+: { sort: 2 } };
)
+ tsCommonPanelOptions
+ tsStandardOptions.withUnit('percentunit');
{
_clusterTemplate:: template.new(
name='cluster',
datasource='$datasource',
query='label_values(node_time_seconds, %s)' % $._config.clusterLabel,
current='',
hide=if $._config.showMultiCluster then '' else '2',
refresh=2,
includeAll=false,
sort=1
),
_clusterVariable::
variable.query.new('cluster')
+ variable.query.withDatasourceFromVariable(datasource)
+ variable.query.queryTypes.withLabelValues(
$._config.clusterLabel,
'node_time_seconds',
)
+ (if $._config.showMultiCluster then variable.query.generalOptions.showOnDashboard.withLabelAndValue() else variable.query.generalOptions.showOnDashboard.withNothing())
+ variable.query.refresh.onTime()
+ variable.query.selectionOptions.withIncludeAll(false)
+ variable.query.withSort(asc=true),
grafanaDashboards+:: {
'node-rsrc-use.json':
dashboard.new(
'%sUSE Method / Node' % $._config.dashboardNamePrefix,
time_from='now-1h',
tags=($._config.dashboardTags),
timezone='utc',
refresh='30s',
graphTooltip='shared_crosshair',
uid=std.md5('node-rsrc-use.json')
)
.addTemplate(datasourceTemplate)
.addTemplate($._clusterTemplate)
.addTemplate(
template.new(
+ dashboard.time.withFrom('now-1h')
+ dashboard.withTags($._config.dashboardTags)
+ dashboard.withTimezone('utc')
+ dashboard.withRefresh('30s')
+ dashboard.graphTooltip.withSharedCrosshair()
+ dashboard.withUid(std.md5('node-rsrc-use.json'))
+ dashboard.withVariables([
datasource,
$._clusterVariable,
variable.query.new('instance')
+ variable.query.withDatasourceFromVariable(datasource)
+ variable.query.queryTypes.withLabelValues(
'instance',
'$datasource',
'label_values(node_exporter_build_info{%(nodeExporterSelector)s, %(clusterLabel)s="$cluster"}, instance)' % $._config,
refresh='time',
sort=1
'node_exporter_build_info{%(nodeExporterSelector)s, %(clusterLabel)s="$cluster"}' % $._config,
)
)
.addRow(
+ variable.query.refresh.onTime()
+ variable.query.withSort(asc=true),
])
+ dashboard.withPanels(
grafana.util.grid.makeGrid([
row.new('CPU')
.addPanel(CPUUtilisation.addTarget(prometheus.target('instance:node_cpu_utilisation:rate%(rateInterval)s{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"} != 0' % $._config, legendFormat='Utilisation')))
.addPanel(CPUSaturation.addTarget(prometheus.target('instance:node_load1_per_cpu:ratio{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"} != 0' % $._config, legendFormat='Saturation')))
)
.addRow(
+ row.withPanels([
CPUUtilisation + tsQueryOptions.withTargets([prometheus.new('$datasource', 'instance:node_cpu_utilisation:rate%(rateInterval)s{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"} != 0' % $._config) + prometheus.withLegendFormat('Utilisation')]),
CPUSaturation + tsQueryOptions.withTargets([prometheus.new('$datasource', 'instance:node_load1_per_cpu:ratio{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"} != 0' % $._config) + prometheus.withLegendFormat('Saturation')]),
]),
row.new('Memory')
.addPanel(memoryUtilisation.addTarget(prometheus.target('instance:node_memory_utilisation:ratio{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"} != 0' % $._config, legendFormat='Utilisation')))
.addPanel(memorySaturation.addTarget(prometheus.target('instance:node_vmstat_pgmajfault:rate%(rateInterval)s{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"} != 0' % $._config, legendFormat='Major page Faults')))
)
.addRow(
+ row.withPanels([
memoryUtilisation + tsQueryOptions.withTargets([prometheus.new('$datasource', 'instance:node_memory_utilisation:ratio{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"} != 0' % $._config) + prometheus.withLegendFormat('Utilisation')]),
memorySaturation + tsQueryOptions.withTargets([prometheus.new('$datasource', 'instance:node_vmstat_pgmajfault:rate%(rateInterval)s{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"} != 0' % $._config) + prometheus.withLegendFormat('Major page Faults')]),
]),
row.new('Network')
.addPanel(
networkUtilisation
.addTarget(prometheus.target('instance:node_network_receive_bytes_excluding_lo:rate%(rateInterval)s{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"} != 0' % $._config, legendFormat='Receive'))
.addTarget(prometheus.target('instance:node_network_transmit_bytes_excluding_lo:rate%(rateInterval)s{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"} != 0' % $._config, legendFormat='Transmit'))
)
.addPanel(
networkSaturation
.addTarget(prometheus.target('instance:node_network_receive_drop_excluding_lo:rate%(rateInterval)s{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"} != 0' % $._config, legendFormat='Receive'))
.addTarget(prometheus.target('instance:node_network_transmit_drop_excluding_lo:rate%(rateInterval)s{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"} != 0' % $._config, legendFormat='Transmit'))
)
)
.addRow(
+ row.withPanels([
networkUtilisation + tsQueryOptions.withTargets([
prometheus.new('$datasource', 'instance:node_network_receive_bytes_excluding_lo:rate%(rateInterval)s{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"} != 0' % $._config) + prometheus.withLegendFormat('Receive'),
prometheus.new('$datasource', 'instance:node_network_transmit_bytes_excluding_lo:rate%(rateInterval)s{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"} != 0' % $._config) + prometheus.withLegendFormat('Transmit'),
]),
networkSaturation + tsQueryOptions.withTargets([
prometheus.new('$datasource', 'instance:node_network_receive_drop_excluding_lo:rate%(rateInterval)s{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"} != 0' % $._config) + prometheus.withLegendFormat('Receive'),
prometheus.new('$datasource', 'instance:node_network_transmit_drop_excluding_lo:rate%(rateInterval)s{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"} != 0' % $._config) + prometheus.withLegendFormat('Transmit'),
]),
]),
row.new('Disk IO')
.addPanel(diskIOUtilisation.addTarget(prometheus.target('instance_device:node_disk_io_time_seconds:rate%(rateInterval)s{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"} != 0' % $._config, legendFormat='{{device}}')))
.addPanel(diskIOSaturation.addTarget(prometheus.target('instance_device:node_disk_io_time_weighted_seconds:rate%(rateInterval)s{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"} != 0' % $._config, legendFormat='{{device}}')))
)
.addRow(
+ row.withPanels([
diskIOUtilisation + tsQueryOptions.withTargets([prometheus.new('$datasource', 'instance_device:node_disk_io_time_seconds:rate%(rateInterval)s{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"} != 0' % $._config) + prometheus.withLegendFormat('{{device}}')]),
diskIOSaturation + tsQueryOptions.withTargets([prometheus.new('$datasource', 'instance_device:node_disk_io_time_weighted_seconds:rate%(rateInterval)s{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"} != 0' % $._config) + prometheus.withLegendFormat('{{device}}')]),
]),
], panelWidth=12, panelHeight=7)
+ grafana.util.grid.makeGrid([
row.new('Disk Space')
.addPanel(
diskSpaceUtilisation.addTarget(prometheus.target(
+ row.withPanels([
diskSpaceUtilisation + tsQueryOptions.withTargets([
prometheus.new(
'$datasource',
|||
sort_desc(1 -
(
@ -204,28 +181,36 @@ local diskSpaceUtilisation =
max without (mountpoint, fstype) (node_filesystem_size_bytes{%(nodeExporterSelector)s, fstype!="", instance="$instance", %(clusterLabel)s="$cluster"})
) != 0
)
||| % $._config, legendFormat='{{device}}'
))
)
||| % $._config
) + prometheus.withLegendFormat('{{device}}'),
]),
]),
], panelWidth=24, panelHeight=7, startY=34),
),
'node-cluster-rsrc-use.json':
dashboard.new(
'%sUSE Method / Cluster' % $._config.dashboardNamePrefix,
time_from='now-1h',
tags=($._config.dashboardTags),
timezone='utc',
refresh='30s',
graphTooltip='shared_crosshair',
uid=std.md5('node-cluster-rsrc-use.json')
)
.addTemplate(datasourceTemplate)
.addTemplate($._clusterTemplate)
.addRow(
+ dashboard.time.withFrom('now-1h')
+ dashboard.withTags($._config.dashboardTags)
+ dashboard.withTimezone('utc')
+ dashboard.withRefresh('30s')
+ dashboard.graphTooltip.withSharedCrosshair()
+ dashboard.withUid(std.md5('node-cluster-rsrc-use.json'))
+ dashboard.withVariables([
datasource,
$._clusterVariable,
variable.query.withDatasourceFromVariable(datasource)
+ variable.query.refresh.onTime()
+ variable.query.withSort(asc=true),
])
+ dashboard.withPanels(
grafana.util.grid.makeGrid([
row.new('CPU')
.addPanel(
CPUUtilisation
.addTarget(prometheus.target(
+ row.withPanels([
CPUUtilisation + tsQueryOptions.withTargets([
prometheus.new(
'$datasource',
|||
((
instance:node_cpu_utilisation:rate%(rateInterval)s{%(nodeExporterSelector)s, %(clusterLabel)s="$cluster"}
@ -233,79 +218,90 @@ local diskSpaceUtilisation =
instance:node_num_cpu:sum{%(nodeExporterSelector)s, %(clusterLabel)s="$cluster"}
) != 0 )
/ scalar(sum(instance:node_num_cpu:sum{%(nodeExporterSelector)s, %(clusterLabel)s="$cluster"}))
||| % $._config, legendFormat='{{ instance }}'
))
)
.addPanel(
CPUSaturation
.addTarget(prometheus.target(
||| % $._config
) + prometheus.withLegendFormat('{{ instance }}'),
]),
CPUSaturation + tsQueryOptions.withTargets([
prometheus.new(
'$datasource',
|||
(
instance:node_load1_per_cpu:ratio{%(nodeExporterSelector)s, %(clusterLabel)s="$cluster"}
/ scalar(count(instance:node_load1_per_cpu:ratio{%(nodeExporterSelector)s, %(clusterLabel)s="$cluster"}))
) != 0
||| % $._config, legendFormat='{{instance}}'
))
)
)
.addRow(
||| % $._config
) + prometheus.withLegendFormat('{{ instance }}'),
]),
]),
row.new('Memory')
.addPanel(
memoryUtilisation
.addTarget(prometheus.target(
+ row.withPanels([
memoryUtilisation + tsQueryOptions.withTargets([
prometheus.new(
'$datasource',
|||
(
instance:node_memory_utilisation:ratio{%(nodeExporterSelector)s, %(clusterLabel)s="$cluster"}
/ scalar(count(instance:node_memory_utilisation:ratio{%(nodeExporterSelector)s, %(clusterLabel)s="$cluster"}))
) != 0
||| % $._config, legendFormat='{{instance}}',
))
)
.addPanel(memorySaturation.addTarget(prometheus.target('instance:node_vmstat_pgmajfault:rate%(rateInterval)s{%(nodeExporterSelector)s, %(clusterLabel)s="$cluster"}' % $._config, legendFormat='{{instance}}')))
)
.addRow(
||| % $._config
) + prometheus.withLegendFormat('{{ instance }}'),
]),
memorySaturation + tsQueryOptions.withTargets([
prometheus.new(
'$datasource',
'instance:node_vmstat_pgmajfault:rate%(rateInterval)s{%(nodeExporterSelector)s, %(clusterLabel)s="$cluster"}' % $._config
) + prometheus.withLegendFormat('{{ instance }}'),
]),
]),
row.new('Network')
.addPanel(
networkUtilisation
.addTarget(prometheus.target('instance:node_network_receive_bytes_excluding_lo:rate%(rateInterval)s{%(nodeExporterSelector)s, %(clusterLabel)s="$cluster"} != 0' % $._config, legendFormat='{{instance}} Receive'))
.addTarget(prometheus.target('instance:node_network_transmit_bytes_excluding_lo:rate%(rateInterval)s{%(nodeExporterSelector)s, %(clusterLabel)s="$cluster"} != 0' % $._config, legendFormat='{{instance}} Transmit'))
)
.addPanel(
networkSaturation
.addTarget(prometheus.target('instance:node_network_receive_drop_excluding_lo:rate%(rateInterval)s{%(nodeExporterSelector)s, %(clusterLabel)s="$cluster"} != 0' % $._config, legendFormat='{{instance}} Receive'))
.addTarget(prometheus.target('instance:node_network_transmit_drop_excluding_lo:rate%(rateInterval)s{%(nodeExporterSelector)s, %(clusterLabel)s="$cluster"} != 0' % $._config, legendFormat='{{instance}} Transmit'))
)
)
.addRow(
+ row.withPanels([
networkUtilisation + tsQueryOptions.withTargets([
prometheus.new(
'$datasource',
'instance:node_network_receive_bytes_excluding_lo:rate%(rateInterval)s{%(nodeExporterSelector)s, %(clusterLabel)s="$cluster"} != 0' % $._config
) + prometheus.withLegendFormat('{{ instance }} Receive'),
prometheus.new(
'$datasource',
'instance:node_network_transmit_bytes_excluding_lo:rate%(rateInterval)s{%(nodeExporterSelector)s, %(clusterLabel)s="$cluster"} != 0' % $._config
) + prometheus.withLegendFormat('{{ instance }} Transmit'),
]),
networkSaturation + tsQueryOptions.withTargets([
prometheus.new(
'$datasource',
'instance:node_network_receive_drop_excluding_lo:rate%(rateInterval)s{%(nodeExporterSelector)s, %(clusterLabel)s="$cluster"} != 0' % $._config
) + prometheus.withLegendFormat('{{ instance }} Receive'),
prometheus.new(
'$datasource',
'instance:node_network_transmit_drop_excluding_lo:rate%(rateInterval)s{%(nodeExporterSelector)s, %(clusterLabel)s="$cluster"} != 0' % $._config
) + prometheus.withLegendFormat('{{ instance }} Transmit'),
]),
]),
row.new('Disk IO')
.addPanel(
diskIOUtilisation
.addTarget(prometheus.target(
+ row.withPanels([
diskIOUtilisation + tsQueryOptions.withTargets([
prometheus.new(
'$datasource',
|||
(
instance_device:node_disk_io_time_seconds:rate%(rateInterval)s{%(nodeExporterSelector)s, %(clusterLabel)s="$cluster"}
/ scalar(count(instance_device:node_disk_io_time_seconds:rate%(rateInterval)s{%(nodeExporterSelector)s, %(clusterLabel)s="$cluster"}))
) != 0
||| % $._config, legendFormat='{{instance}} {{device}}'
))
)
.addPanel(
diskIOSaturation
.addTarget(prometheus.target(
||| % $._config
) + prometheus.withLegendFormat('{{ instance }} {{device}}'),
]),
diskIOSaturation + tsQueryOptions.withTargets([prometheus.new(
'$datasource',
|||
(
instance_device:node_disk_io_time_weighted_seconds:rate%(rateInterval)s{%(nodeExporterSelector)s, %(clusterLabel)s="$cluster"}
/ scalar(count(instance_device:node_disk_io_time_weighted_seconds:rate%(rateInterval)s{%(nodeExporterSelector)s, %(clusterLabel)s="$cluster"}))
) != 0
||| % $._config, legendFormat='{{instance}} {{device}}'
))
)
)
.addRow(
||| % $._config
) + prometheus.withLegendFormat('{{ instance }} {{device}}')]),
]),
], panelWidth=12, panelHeight=7)
+ grafana.util.grid.makeGrid([
row.new('Disk Space')
.addPanel(
diskSpaceUtilisation
.addTarget(prometheus.target(
+ row.withPanels([
diskSpaceUtilisation + tsQueryOptions.withTargets([
prometheus.new(
'$datasource',
|||
sum without (device) (
max without (fstype, mountpoint) ((
@ -315,28 +311,37 @@ local diskSpaceUtilisation =
) != 0)
)
/ scalar(sum(max without (fstype, mountpoint) (node_filesystem_size_bytes{%(nodeExporterSelector)s, %(fsSelector)s, %(fsMountpointSelector)s, %(clusterLabel)s="$cluster"})))
||| % $._config, legendFormat='{{instance}}'
))
)
||| % $._config
) + prometheus.withLegendFormat('{{ instance }}'),
]),
]),
], panelWidth=24, panelHeight=7, startY=34),
),
} +
if $._config.showMultiCluster then {
'node-multicluster-rsrc-use.json':
dashboard.new(
'%sUSE Method / Multi-cluster' % $._config.dashboardNamePrefix,
time_from='now-1h',
tags=($._config.dashboardTags),
timezone='utc',
refresh='30s',
graphTooltip='shared_crosshair',
uid=std.md5('node-multicluster-rsrc-use.json')
)
.addTemplate(datasourceTemplate)
.addRow(
+ dashboard.time.withFrom('now-1h')
+ dashboard.withTags($._config.dashboardTags)
+ dashboard.withTimezone('utc')
+ dashboard.withRefresh('30s')
+ dashboard.graphTooltip.withSharedCrosshair()
+ dashboard.withUid(std.md5('node-multicluster-rsrc-use.json'))
+ dashboard.withVariables([
datasource,
variable.query.withDatasourceFromVariable(datasource)
+ variable.query.refresh.onTime()
+ variable.query.withSort(asc=true),
])
+ dashboard.withPanels(
grafana.util.grid.makeGrid([
row.new('CPU')
.addPanel(
CPUUtilisation
.addTarget(prometheus.target(
+ row.withPanels([
CPUUtilisation + tsQueryOptions.withTargets([
prometheus.new(
'$datasource',
|||
sum(
((
@ -346,112 +351,116 @@ local diskSpaceUtilisation =
) != 0)
/ scalar(sum(instance:node_num_cpu:sum{%(nodeExporterSelector)s}))
) by (%(clusterLabel)s)
||| % $._config, legendFormat='{{%(clusterLabel)s}}' % $._config
))
)
.addPanel(
CPUSaturation
.addTarget(prometheus.target(
||| % $._config
) + prometheus.withLegendFormat('{{%(clusterLabel)s}}'),
]),
CPUSaturation + tsQueryOptions.withTargets([
prometheus.new(
'$datasource',
|||
sum((
instance:node_load1_per_cpu:ratio{%(nodeExporterSelector)s}
/ scalar(count(instance:node_load1_per_cpu:ratio{%(nodeExporterSelector)s}))
) != 0) by (%(clusterLabel)s)
||| % $._config, legendFormat='{{%(clusterLabel)s}}' % $._config
))
)
)
.addRow(
||| % $._config
) + prometheus.withLegendFormat('{{%(clusterLabel)s}}'),
]),
]),
row.new('Memory')
.addPanel(
memoryUtilisation
.addTarget(prometheus.target(
+ row.withPanels([
memoryUtilisation + tsQueryOptions.withTargets([
prometheus.new(
'$datasource',
|||
sum((
instance:node_memory_utilisation:ratio{%(nodeExporterSelector)s}
/ scalar(count(instance:node_memory_utilisation:ratio{%(nodeExporterSelector)s}))
) != 0) by (%(clusterLabel)s)
||| % $._config, legendFormat='{{%(clusterLabel)s}}' % $._config
))
)
.addPanel(
memorySaturation
.addTarget(prometheus.target(
||| % $._config
) + prometheus.withLegendFormat('{{%(clusterLabel)s}}'),
]),
memorySaturation + tsQueryOptions.withTargets([
prometheus.new(
'$datasource',
|||
sum((
instance:node_vmstat_pgmajfault:rate%(rateInterval)s{%(nodeExporterSelector)s}
) != 0) by (%(clusterLabel)s)
||| % $._config, legendFormat='{{%(clusterLabel)s}}' % $._config
))
)
)
.addRow(
|||
% $._config
) + prometheus.withLegendFormat('{{%(clusterLabel)s}}'),
]),
]),
row.new('Network')
.addPanel(
networkUtilisation
.addTarget(prometheus.target(
+ row.withPanels([
networkUtilisation + tsQueryOptions.withTargets([
prometheus.new(
'$datasource',
|||
sum((
instance:node_network_receive_bytes_excluding_lo:rate%(rateInterval)s{%(nodeExporterSelector)s}
) != 0) by (%(clusterLabel)s)
||| % $._config, legendFormat='{{%(clusterLabel)s}} Receive' % $._config
))
.addTarget(prometheus.target(
||| % $._config
) + prometheus.withLegendFormat('{{%(clusterLabel)s}} Receive'),
prometheus.new(
'$datasource',
|||
sum((
instance:node_network_transmit_bytes_excluding_lo:rate%(rateInterval)s{%(nodeExporterSelector)s}
) != 0) by (%(clusterLabel)s)
||| % $._config, legendFormat='{{%(clusterLabel)s}} Transmit' % $._config
))
)
.addPanel(
networkSaturation
.addTarget(prometheus.target(
||| % $._config
) + prometheus.withLegendFormat('{{%(clusterLabel)s}} Transmit'),
]),
networkSaturation + tsQueryOptions.withTargets([
prometheus.new(
'$datasource',
|||
sum((
instance:node_network_receive_drop_excluding_lo:rate%(rateInterval)s{%(nodeExporterSelector)s}
) != 0) by (%(clusterLabel)s)
||| % $._config, legendFormat='{{%(clusterLabel)s}} Receive' % $._config
))
.addTarget(prometheus.target(
||| % $._config
) + prometheus.withLegendFormat('{{%(clusterLabel)s}} Receive'),
prometheus.new(
'$datasource',
|||
sum((
instance:node_network_transmit_drop_excluding_lo:rate%(rateInterval)s{%(nodeExporterSelector)s}
) != 0) by (%(clusterLabel)s)
||| % $._config, legendFormat='{{%(clusterLabel)s}} Transmit' % $._config
))
)
)
.addRow(
||| % $._config
) + prometheus.withLegendFormat('{{%(clusterLabel)s}} Transmit'),
]),
]),
row.new('Disk IO')
.addPanel(
diskIOUtilisation
.addTarget(prometheus.target(
+ row.withPanels([
diskIOUtilisation + tsQueryOptions.withTargets([
prometheus.new(
'$datasource',
|||
sum((
instance_device:node_disk_io_time_seconds:rate%(rateInterval)s{%(nodeExporterSelector)s}
/ scalar(count(instance_device:node_disk_io_time_seconds:rate%(rateInterval)s{%(nodeExporterSelector)s}))
) != 0) by (%(clusterLabel)s, device)
||| % $._config, legendFormat='{{%(clusterLabel)s}} {{device}}' % $._config
))
)
.addPanel(
diskIOSaturation
.addTarget(prometheus.target(
||| % $._config
) + prometheus.withLegendFormat('{{%(clusterLabel)s}} {{device}}'),
]),
diskIOSaturation + tsQueryOptions.withTargets([prometheus.new(
'$datasource',
|||
sum((
instance_device:node_disk_io_time_weighted_seconds:rate%(rateInterval)s{%(nodeExporterSelector)s}
/ scalar(count(instance_device:node_disk_io_time_weighted_seconds:rate%(rateInterval)s{%(nodeExporterSelector)s}))
) != 0) by (%(clusterLabel)s, device)
||| % $._config, legendFormat='{{%(clusterLabel)s}} {{device}}' % $._config
))
)
)
.addRow(
||| % $._config
) + prometheus.withLegendFormat('{{%(clusterLabel)s}} {{device}}')]),
]),
], panelWidth=12, panelHeight=7)
+ grafana.util.grid.makeGrid([
row.new('Disk Space')
.addPanel(
diskSpaceUtilisation
.addTarget(prometheus.target(
+ row.withPanels([
diskSpaceUtilisation + tsQueryOptions.withTargets([
prometheus.new(
'$datasource',
|||
sum (
sum without (device) (
@ -461,9 +470,11 @@ local diskSpaceUtilisation =
)
/ scalar(sum(max without (fstype, mountpoint) (node_filesystem_size_bytes{%(nodeExporterSelector)s, %(fsSelector)s, %(fsMountpointSelector)s})))
) by (%(clusterLabel)s)
||| % $._config, legendFormat='{{%(clusterLabel)s}}' % $._config
))
)
||| % $._config
) + prometheus.withLegendFormat('{{%(clusterLabel)s}}'),
]),
]),
], panelWidth=24, panelHeight=7, startY=34),
),
} else {},
}

View file

@ -4,20 +4,11 @@
{
"source": {
"git": {
"remote": "https://github.com/grafana/grafonnet-lib.git",
"subdir": "grafonnet"
"remote": "https://github.com/grafana/grafonnet.git",
"subdir": "gen/grafonnet-latest"
}
},
"version": "master"
},
{
"source": {
"git": {
"remote": "https://github.com/grafana/grafonnet-lib.git",
"subdir": "grafonnet-7.0"
}
},
"version": "master"
"version": "main"
}
],
"legacyImports": false

View file

@ -1,76 +1,85 @@
local grafana = import 'github.com/grafana/grafonnet-lib/grafonnet/grafana.libsonnet';
local grafana = import 'github.com/grafana/grafonnet/gen/grafonnet-latest/main.libsonnet';
local dashboard = grafana.dashboard;
local row = grafana.row;
local prometheus = grafana.prometheus;
local template = grafana.template;
local graphPanel = grafana.graphPanel;
local grafana70 = import 'github.com/grafana/grafonnet-lib/grafonnet-7.0/grafana.libsonnet';
local gaugePanel = grafana70.panel.gauge;
local table = grafana70.panel.table;
local row = grafana.panel.row;
local prometheus = grafana.query.prometheus;
local variable = dashboard.variable;
local timeSeriesPanel = grafana.panel.timeSeries;
local tsOptions = timeSeriesPanel.options;
local tsStandardOptions = timeSeriesPanel.standardOptions;
local tsQueryOptions = timeSeriesPanel.queryOptions;
local tsCustom = timeSeriesPanel.fieldConfig.defaults.custom;
local gaugePanel = grafana.panel.gauge;
local gaugeStep = gaugePanel.standardOptions.threshold.step;
local table = grafana.panel.table;
local tableStep = table.standardOptions.threshold.step;
local tableOverride = table.standardOptions.override;
local tableTransformation = table.queryOptions.transformation;
{
new(config=null, platform=null, uid=null):: {
local prometheusDatasourceTemplate = {
current: {
text: 'default',
value: 'default',
},
hide: 0,
label: 'Data Source',
name: 'datasource',
options: [],
query: 'prometheus',
refresh: 1,
regex: '',
type: 'datasource',
},
local clusterTemplatePrototype =
template.new(
'cluster',
'$datasource',
'',
hide=if config.showMultiCluster then '' else '2',
refresh='time',
label='Cluster',
local prometheusDatasourceVariable = variable.datasource.new(
'datasource', 'prometheus'
),
local clusterTemplate =
if platform == 'Darwin' then
clusterTemplatePrototype
{ query: 'label_values(node_uname_info{%(nodeExporterSelector)s, sysname="Darwin"}, %(clusterLabel)s)' % config }
else
clusterTemplatePrototype
{ query: 'label_values(node_uname_info{%(nodeExporterSelector)s, sysname!="Darwin"}, %(clusterLabel)s)' % config },
local instanceTemplatePrototype =
template.new(
local clusterVariablePrototype =
variable.query.new('cluster')
+ variable.query.withDatasourceFromVariable(prometheusDatasourceVariable)
+ (if config.showMultiCluster then variable.query.generalOptions.showOnDashboard.withLabelAndValue() else variable.query.generalOptions.showOnDashboard.withNothing())
+ variable.query.refresh.onTime()
+ variable.query.generalOptions.withLabel('Cluster'),
local clusterVariable =
if platform == 'Darwin' then
clusterVariablePrototype
+ variable.query.queryTypes.withLabelValues(
' %(clusterLabel)s' % config,
'node_uname_info{%(nodeExporterSelector)s, sysname="Darwin"}' % config,
)
else
clusterVariablePrototype
+ variable.query.queryTypes.withLabelValues(
'%(clusterLabel)s' % config,
'node_uname_info{%(nodeExporterSelector)s, sysname!="Darwin"}' % config,
),
local instanceVariablePrototype =
variable.query.new('instance')
+ variable.query.withDatasourceFromVariable(prometheusDatasourceVariable)
+ variable.query.refresh.onTime()
+ variable.query.generalOptions.withLabel('Instance'),
local instanceVariable =
if platform == 'Darwin' then
instanceVariablePrototype
+ variable.query.queryTypes.withLabelValues(
'instance',
'$datasource',
'',
refresh='time',
label='Instance',
),
local instanceTemplate =
if platform == 'Darwin' then
instanceTemplatePrototype
{ query: 'label_values(node_uname_info{%(nodeExporterSelector)s, %(clusterLabel)s="$cluster", sysname="Darwin"}, instance)' % config }
'node_uname_info{%(nodeExporterSelector)s, %(clusterLabel)s="$cluster", sysname="Darwin"}' % config,
)
else
instanceTemplatePrototype
{ query: 'label_values(node_uname_info{%(nodeExporterSelector)s, %(clusterLabel)s="$cluster", sysname!="Darwin"}, instance)' % config },
instanceVariablePrototype
+ variable.query.queryTypes.withLabelValues(
'instance',
'node_uname_info{%(nodeExporterSelector)s, %(clusterLabel)s="$cluster", sysname!="Darwin"}' % config,
),
local idleCPU =
graphPanel.new(
'CPU Usage',
datasource='$datasource',
span=6,
format='percentunit',
max=1,
min=0,
stack=true,
)
.addTarget(prometheus.target(
timeSeriesPanel.new('CPU Usage')
+ variable.query.withDatasourceFromVariable(prometheusDatasourceVariable)
+ tsStandardOptions.withUnit('percentunit')
+ tsCustom.stacking.withMode('normal')
+ tsStandardOptions.withMax(1)
+ tsStandardOptions.withMin(0)
+ tsOptions.tooltip.withMode('multi')
+ tsCustom.withFillOpacity(10)
+ tsCustom.withShowPoints('never')
+ tsQueryOptions.withTargets([
prometheus.new(
'$datasource',
|||
(
(1 - sum without (mode) (rate(node_cpu_seconds_total{%(nodeExporterSelector)s, mode=~"idle|iowait|steal", instance="$instance", %(clusterLabel)s="$cluster"}[$__rate_interval])))
@ -78,36 +87,42 @@ local table = grafana70.panel.table;
count without (cpu, mode) (node_cpu_seconds_total{%(nodeExporterSelector)s, mode="idle", instance="$instance", %(clusterLabel)s="$cluster"})
)
||| % config,
legendFormat='{{cpu}}',
intervalFactor=5,
)),
)
+ prometheus.withLegendFormat('{{cpu}}')
+ prometheus.withIntervalFactor(5),
]),
local systemLoad =
graphPanel.new(
'Load Average',
datasource='$datasource',
span=6,
format='short',
min=0,
fill=0,
)
.addTarget(prometheus.target('node_load1{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"}' % config, legendFormat='1m load average'))
.addTarget(prometheus.target('node_load5{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"}' % config, legendFormat='5m load average'))
.addTarget(prometheus.target('node_load15{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"}' % config, legendFormat='15m load average'))
.addTarget(prometheus.target('count(node_cpu_seconds_total{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster", mode="idle"})' % config, legendFormat='logical cores')),
timeSeriesPanel.new('Load Average')
+ variable.query.withDatasourceFromVariable(prometheusDatasourceVariable)
+ tsStandardOptions.withUnit('short')
+ tsStandardOptions.withMin(0)
+ tsCustom.withFillOpacity(0)
+ tsCustom.withShowPoints('never')
+ tsOptions.tooltip.withMode('multi')
+ tsQueryOptions.withTargets([
prometheus.new('$datasource', 'node_load1{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"}' % config) + prometheus.withLegendFormat('1m load average'),
prometheus.new('$datasource', 'node_load5{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"}' % config) + prometheus.withLegendFormat('5m load average'),
prometheus.new('$datasource', 'node_load15{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"}' % config) + prometheus.withLegendFormat('15m load average'),
prometheus.new('$datasource', 'count(node_cpu_seconds_total{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster", mode="idle"})' % config) + prometheus.withLegendFormat('logical cores'),
]),
local memoryGraphPanelPrototype =
graphPanel.new(
'Memory Usage',
datasource='$datasource',
span=9,
format='bytes',
min=0,
),
timeSeriesPanel.new('Memory Usage')
+ variable.query.withDatasourceFromVariable(prometheusDatasourceVariable)
+ tsStandardOptions.withUnit('bytes')
+ tsStandardOptions.withMin(0)
+ tsOptions.tooltip.withMode('multi')
+ tsCustom.withFillOpacity(10)
+ tsCustom.withShowPoints('never'),
local memoryGraph =
if platform == 'Linux' then
memoryGraphPanelPrototype { stack: true }
.addTarget(prometheus.target(
memoryGraphPanelPrototype
+ tsCustom.stacking.withMode('normal')
+ tsQueryOptions.withTargets([
prometheus.new(
'$datasource',
|||
(
node_memory_MemTotal_bytes{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"}
@ -119,16 +134,19 @@ local table = grafana70.panel.table;
node_memory_Cached_bytes{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"}
)
||| % config,
legendFormat='memory used'
))
.addTarget(prometheus.target('node_memory_Buffers_bytes{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"}' % config, legendFormat='memory buffers'))
.addTarget(prometheus.target('node_memory_Cached_bytes{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"}' % config, legendFormat='memory cached'))
.addTarget(prometheus.target('node_memory_MemFree_bytes{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"}' % config, legendFormat='memory free'))
) + prometheus.withLegendFormat('memory used'),
prometheus.new('$datasource', 'node_memory_Buffers_bytes{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"}' % config) + prometheus.withLegendFormat('memory buffers'),
prometheus.new('$datasource', 'node_memory_Cached_bytes{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"}' % config) + prometheus.withLegendFormat('memory cached'),
prometheus.new('$datasource', 'node_memory_MemFree_bytes{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"}' % config) + prometheus.withLegendFormat('memory free'),
])
else if platform == 'Darwin' then
// not useful to stack
memoryGraphPanelPrototype { stack: false }
.addTarget(prometheus.target('node_memory_total_bytes{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"}' % config, legendFormat='Physical Memory'))
.addTarget(prometheus.target(
memoryGraphPanelPrototype
+ tsCustom.stacking.withMode('none')
+ tsQueryOptions.withTargets([
prometheus.new('$datasource', 'node_memory_total_bytes{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"}' % config) + prometheus.withLegendFormat('Physical Memory'),
prometheus.new(
'$datasource',
|||
(
node_memory_internal_bytes{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"} -
@ -136,50 +154,61 @@ local table = grafana70.panel.table;
node_memory_wired_bytes{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"} +
node_memory_compressed_bytes{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"}
)
||| % config, legendFormat='Memory Used'
))
.addTarget(prometheus.target(
||| % config
) + prometheus.withLegendFormat(
'Memory Used'
),
prometheus.new(
'$datasource',
|||
(
node_memory_internal_bytes{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"} -
node_memory_purgeable_bytes{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"}
)
||| % config, legendFormat='App Memory'
))
.addTarget(prometheus.target('node_memory_wired_bytes{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"}' % config, legendFormat='Wired Memory'))
.addTarget(prometheus.target('node_memory_compressed_bytes{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"}' % config, legendFormat='Compressed'))
||| % config
) + prometheus.withLegendFormat(
'App Memory'
),
prometheus.new('$datasource', 'node_memory_wired_bytes{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"}' % config) + prometheus.withLegendFormat('Wired Memory'),
prometheus.new('$datasource', 'node_memory_compressed_bytes{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"}' % config) + prometheus.withLegendFormat('Compressed'),
])
else if platform == 'AIX' then
memoryGraphPanelPrototype { stack: false }
.addTarget(prometheus.target('node_memory_total_bytes{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"}' % config, legendFormat='Physical Memory'))
.addTarget(prometheus.target(
memoryGraphPanelPrototype
+ tsCustom.stacking.withMode('none')
+ tsQueryOptions.withTargets([
prometheus.new('$datasource', 'node_memory_total_bytes{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"}' % config) + prometheus.withLegendFormat('Physical Memory'),
prometheus.new(
'$datasource',
|||
(
node_memory_total_bytes{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"} -
node_memory_available_bytes{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster"}
)
||| % config, legendFormat='Memory Used'
)),
||| % config
) + prometheus.withLegendFormat('Memory Used'),
]),
// NOTE: avg() is used to circumvent a label change caused by a node_exporter rollout.
local memoryGaugePanelPrototype =
gaugePanel.new(
title='Memory Usage',
datasource='$datasource',
)
.addThresholdStep('rgba(50, 172, 45, 0.97)')
.addThresholdStep('rgba(237, 129, 40, 0.89)', 80)
.addThresholdStep('rgba(245, 54, 54, 0.9)', 90)
.setFieldConfig(max=100, min=0, unit='percent')
+ {
span: 3,
},
gaugePanel.new('Memory Usage')
+ variable.query.withDatasourceFromVariable(prometheusDatasourceVariable)
+ gaugePanel.standardOptions.thresholds.withSteps([
gaugeStep.withColor('rgba(50, 172, 45, 0.97)'),
gaugeStep.withColor('rgba(237, 129, 40, 0.89)') + gaugeStep.withValue(80),
gaugeStep.withColor('rgba(245, 54, 54, 0.9)') + gaugeStep.withValue(90),
])
+ gaugePanel.standardOptions.withMax(100)
+ gaugePanel.standardOptions.withMin(0)
+ gaugePanel.standardOptions.withUnit('percent'),
local memoryGauge =
if platform == 'Linux' then
memoryGaugePanelPrototype
.addTarget(prometheus.target(
+ gaugePanel.queryOptions.withTargets([
prometheus.new(
'$datasource',
|||
100 -
(
@ -188,11 +217,14 @@ local table = grafana70.panel.table;
* 100
)
||| % config,
))
),
])
else if platform == 'Darwin' then
memoryGaugePanelPrototype
.addTarget(prometheus.target(
+ gaugePanel.queryOptions.withTargets([
prometheus.new(
'$datasource',
|||
(
(
@ -206,10 +238,14 @@ local table = grafana70.panel.table;
*
100
||| % config
))
),
])
else if platform == 'AIX' then
memoryGaugePanelPrototype
.addTarget(prometheus.target(
+ gaugePanel.queryOptions.withTargets([
prometheus.new(
'$datasource',
|||
100 -
(
@ -218,156 +254,94 @@ local table = grafana70.panel.table;
* 100
)
||| % config
)),
),
]),
local diskIO =
graphPanel.new(
'Disk I/O',
datasource='$datasource',
span=6,
min=0,
fill=0,
)
timeSeriesPanel.new('Disk I/O')
+ variable.query.withDatasourceFromVariable(prometheusDatasourceVariable)
+ tsStandardOptions.withMin(0)
+ tsCustom.withFillOpacity(0)
+ tsCustom.withShowPoints('never')
+ tsOptions.tooltip.withMode('multi')
+ tsQueryOptions.withTargets([
// TODO: Does it make sense to have those three in the same panel?
.addTarget(prometheus.target(
'rate(node_disk_read_bytes_total{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster", %(diskDeviceSelector)s}[$__rate_interval])' % config,
legendFormat='{{device}} read',
intervalFactor=1,
))
.addTarget(prometheus.target(
'rate(node_disk_written_bytes_total{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster", %(diskDeviceSelector)s}[$__rate_interval])' % config,
legendFormat='{{device}} written',
intervalFactor=1,
))
.addTarget(prometheus.target(
'rate(node_disk_io_time_seconds_total{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster", %(diskDeviceSelector)s}[$__rate_interval])' % config,
legendFormat='{{device}} io time',
intervalFactor=1,
)) +
{
seriesOverrides: [
{
alias: '/ read| written/',
yaxis: 1,
},
{
alias: '/ io time/',
yaxis: 2,
},
],
yaxes: [
self.yaxe(format='Bps'),
self.yaxe(format='percentunit'),
],
},
prometheus.new('$datasource', 'rate(node_disk_read_bytes_total{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster", %(diskDeviceSelector)s}[$__rate_interval])' % config)
+ prometheus.withLegendFormat('{{device}} read')
+ prometheus.withIntervalFactor(1),
prometheus.new('$datasource', 'rate(node_disk_written_bytes_total{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster", %(diskDeviceSelector)s}[$__rate_interval])' % config)
+ prometheus.withLegendFormat('{{device}} written')
+ prometheus.withIntervalFactor(1),
prometheus.new('$datasource', 'rate(node_disk_io_time_seconds_total{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster", %(diskDeviceSelector)s}[$__rate_interval])' % config)
+ prometheus.withLegendFormat('{{device}} io time')
+ prometheus.withIntervalFactor(1),
])
+ tsStandardOptions.withOverrides(
[
tsStandardOptions.override.byRegexp.new('/ read| written/')
+ tsStandardOptions.override.byRegexp.withPropertiesFromOptions(
tsStandardOptions.withUnit('Bps')
),
tsStandardOptions.override.byRegexp.new('/ io time/')
+ tsStandardOptions.override.byRegexp.withPropertiesFromOptions(tsStandardOptions.withUnit('percentunit')),
]
),
local diskSpaceUsage =
table.new(
title='Disk Space Usage',
datasource='$datasource',
)
.setFieldConfig(unit='decbytes')
.addThresholdStep(color='green', value=null)
.addThresholdStep(color='yellow', value=0.8)
.addThresholdStep(color='red', value=0.9)
.addTarget(prometheus.target(
|||
max by (mountpoint) (node_filesystem_size_bytes{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster", %(fsSelector)s, %(fsMountpointSelector)s})
||| % config,
legendFormat='',
instant=true,
format='table'
))
.addTarget(prometheus.target(
|||
max by (mountpoint) (node_filesystem_avail_bytes{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster", %(fsSelector)s, %(fsMountpointSelector)s})
||| % config,
legendFormat='',
instant=true,
format='table'
))
.addOverride(
matcher={
id: 'byName',
options: 'Mounted on',
},
properties=[
{
id: 'custom.width',
value: 260,
},
],
)
.addOverride(
matcher={
id: 'byName',
options: 'Size',
},
properties=[
{
id: 'custom.width',
value: 93,
},
],
)
.addOverride(
matcher={
id: 'byName',
options: 'Used',
},
properties=[
{
id: 'custom.width',
value: 72,
},
],
)
.addOverride(
matcher={
id: 'byName',
options: 'Available',
},
properties=[
{
id: 'custom.width',
value: 88,
},
],
)
.addOverride(
matcher={
id: 'byName',
options: 'Used, %',
},
properties=[
{
id: 'unit',
value: 'percentunit',
},
{
id: 'custom.displayMode',
value: 'gradient-gauge',
},
{
id: 'max',
value: 1,
},
{
id: 'min',
value: 0,
},
table.new('Disk Space Usage')
+ variable.query.withDatasourceFromVariable(prometheusDatasourceVariable)
+ table.standardOptions.withUnit('decbytes')
+ table.standardOptions.thresholds.withSteps(
[
tableStep.withColor('green'),
tableStep.withColor('yellow') + gaugeStep.withValue(0.8),
tableStep.withColor('red') + gaugeStep.withValue(0.9),
]
)
+ { span: 6 }
+ {
transformations: [
+ table.queryOptions.withTargets([
prometheus.new(
'$datasource',
|||
max by (mountpoint) (node_filesystem_size_bytes{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster", %(fsSelector)s, %(fsMountpointSelector)s})
||| % config
)
+ prometheus.withLegendFormat('')
+ prometheus.withInstant()
+ prometheus.withFormat('table'),
prometheus.new(
'$datasource',
|||
max by (mountpoint) (node_filesystem_avail_bytes{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster", %(fsSelector)s, %(fsMountpointSelector)s})
||| % config
)
+ prometheus.withLegendFormat('')
+ prometheus.withInstant()
+ prometheus.withFormat('table'),
])
+ table.standardOptions.withOverrides([
tableOverride.byName.new('Mounted on')
+ tableOverride.byName.withProperty('custom.width', 260),
tableOverride.byName.new('Size')
+ tableOverride.byName.withProperty('custom.width', 93),
tableOverride.byName.new('Used')
+ tableOverride.byName.withProperty('custom.width', 72),
tableOverride.byName.new('Available')
+ tableOverride.byName.withProperty('custom.width', 88),
tableOverride.byName.new('Used, %')
+ tableOverride.byName.withProperty('unit', 'percentunit')
+ tableOverride.byName.withPropertiesFromOptions(
table.fieldConfig.defaults.custom.withCellOptions(
{ type: 'gauge' },
)
)
+ tableOverride.byName.withProperty('max', 1)
+ tableOverride.byName.withProperty('min', 0),
])
+ table.queryOptions.withTransformations([
tableTransformation.withId('groupBy')
+ tableTransformation.withOptions(
{
id: 'groupBy',
options: {
fields: {
'Value #A': {
aggregations: [
@ -386,15 +360,12 @@ local table = grafana70.panel.table;
operation: 'groupby',
},
},
},
},
}
),
tableTransformation.withId('merge'),
tableTransformation.withId('calculateField')
+ tableTransformation.withOptions(
{
id: 'merge',
options: {},
},
{
id: 'calculateField',
options: {
alias: 'Used',
binary: {
left: 'Value #A (lastNotNull)',
@ -406,11 +377,11 @@ local table = grafana70.panel.table;
reduce: {
reducer: 'sum',
},
},
},
}
),
tableTransformation.withId('calculateField')
+ tableTransformation.withOptions(
{
id: 'calculateField',
options: {
alias: 'Used, %',
binary: {
left: 'Used',
@ -422,11 +393,11 @@ local table = grafana70.panel.table;
reduce: {
reducer: 'sum',
},
},
},
}
),
tableTransformation.withId('organize')
+ tableTransformation.withOptions(
{
id: 'organize',
options: {
excludeByName: {},
indexByName: {},
renameByName: {
@ -434,127 +405,131 @@ local table = grafana70.panel.table;
'Value #B (lastNotNull)': 'Available',
mountpoint: 'Mounted on',
},
},
},
}
),
tableTransformation.withId('sortBy')
+ tableTransformation.withOptions(
{
id: 'sortBy',
options: {
fields: {},
sort: [
{
field: 'Mounted on',
},
],
},
},
],
},
}
),
]),
local networkReceived =
graphPanel.new(
'Network Received',
description='Network received (bits/s)',
datasource='$datasource',
span=6,
format='bps',
min=0,
fill=0,
)
.addTarget(prometheus.target(
'rate(node_network_receive_bytes_total{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster", device!="lo"}[$__rate_interval]) * 8' % config,
legendFormat='{{device}}',
intervalFactor=1,
)),
timeSeriesPanel.new('Network Received')
+ timeSeriesPanel.panelOptions.withDescription('Network received (bits/s)')
+ variable.query.withDatasourceFromVariable(prometheusDatasourceVariable)
+ tsStandardOptions.withUnit('bps')
+ tsStandardOptions.withMin(0)
+ tsCustom.withFillOpacity(0)
+ tsCustom.withShowPoints('never')
+ tsOptions.tooltip.withMode('multi')
+ tsQueryOptions.withTargets([
prometheus.new('$datasource', 'rate(node_network_receive_bytes_total{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster", device!="lo"}[$__rate_interval]) * 8' % config)
+ prometheus.withLegendFormat('{{device}}')
+ prometheus.withIntervalFactor(1),
]),
local networkTransmitted =
graphPanel.new(
'Network Transmitted',
description='Network transmitted (bits/s)',
datasource='$datasource',
span=6,
format='bps',
min=0,
fill=0,
)
.addTarget(prometheus.target(
'rate(node_network_transmit_bytes_total{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster", device!="lo"}[$__rate_interval]) * 8' % config,
legendFormat='{{device}}',
intervalFactor=1,
)),
timeSeriesPanel.new('Network Transmitted')
+ timeSeriesPanel.panelOptions.withDescription('Network transmitted (bits/s)')
+ variable.query.withDatasourceFromVariable(prometheusDatasourceVariable)
+ tsStandardOptions.withUnit('bps')
+ tsStandardOptions.withMin(0)
+ tsCustom.withFillOpacity(0)
+ tsOptions.tooltip.withMode('multi')
+ tsQueryOptions.withTargets([
prometheus.new('$datasource', 'rate(node_network_transmit_bytes_total{%(nodeExporterSelector)s, instance="$instance", %(clusterLabel)s="$cluster", device!="lo"}[$__rate_interval]) * 8' % config)
+ prometheus.withLegendFormat('{{device}}')
+ prometheus.withIntervalFactor(1),
]),
local cpuRow =
row.new('CPU')
.addPanel(idleCPU)
.addPanel(systemLoad),
+ row.withPanels([
idleCPU,
systemLoad,
]),
local memoryRow =
row.new('Memory')
.addPanel(memoryGraph)
.addPanel(memoryGauge),
local memoryRow = [
row.new('Memory') + row.gridPos.withY(8),
memoryGraph + row.gridPos.withX(0) + row.gridPos.withY(9) + row.gridPos.withH(7) + row.gridPos.withW(18),
memoryGauge + row.gridPos.withX(18) + row.gridPos.withY(9) + row.gridPos.withH(7) + row.gridPos.withW(6),
],
local diskRow =
row.new('Disk')
.addPanel(diskIO)
.addPanel(diskSpaceUsage),
+ row.withPanels([
diskIO,
diskSpaceUsage,
]),
local networkRow =
row.new('Network')
.addPanel(networkReceived)
.addPanel(networkTransmitted),
+ row.withPanels([
networkReceived,
networkTransmitted,
]),
local rows =
[
local panels =
grafana.util.grid.makeGrid([
cpuRow,
memoryRow,
], panelWidth=12, panelHeight=7)
+ memoryRow
+ grafana.util.grid.makeGrid([
diskRow,
networkRow,
],
], panelWidth=12, panelHeight=7, startY=18),
local templates =
local variables =
[
prometheusDatasourceTemplate,
clusterTemplate,
instanceTemplate,
prometheusDatasourceVariable,
clusterVariable,
instanceVariable,
],
dashboard: if platform == 'Linux' then
dashboard.new(
'%sNodes' % config.dashboardNamePrefix,
time_from='now-1h',
tags=(config.dashboardTags),
timezone='utc',
refresh='30s',
uid=std.md5(uid),
graphTooltip='shared_crosshair'
)
.addTemplates(templates)
.addRows(rows)
+ dashboard.time.withFrom('now-1h')
+ dashboard.withTags(config.dashboardTags)
+ dashboard.withTimezone('utc')
+ dashboard.withRefresh('30s')
+ dashboard.withUid(std.md5(uid))
+ dashboard.graphTooltip.withSharedCrosshair()
+ dashboard.withVariables(variables)
+ dashboard.withPanels(panels)
else if platform == 'Darwin' then
dashboard.new(
'%sMacOS' % config.dashboardNamePrefix,
time_from='now-1h',
tags=(config.dashboardTags),
timezone='utc',
refresh='30s',
uid=std.md5(uid),
graphTooltip='shared_crosshair'
)
.addTemplates(templates)
.addRows(rows)
+ dashboard.time.withFrom('now-1h')
+ dashboard.withTags(config.dashboardTags)
+ dashboard.withTimezone('utc')
+ dashboard.withRefresh('30s')
+ dashboard.withUid(std.md5(uid))
+ dashboard.graphTooltip.withSharedCrosshair()
+ dashboard.withVariables(variables)
+ dashboard.withPanels(panels)
else if platform == 'AIX' then
dashboard.new(
'%sAIX' % config.dashboardNamePrefix,
time_from='now-1h',
tags=(config.dashboardTags),
timezone='utc',
refresh='30s',
uid=std.md5(uid),
graphTooltip='shared_crosshair'
)
.addTemplates(templates)
.addRows(rows),
+ dashboard.time.withFrom('now-1h')
+ dashboard.withTags(config.dashboardTags)
+ dashboard.withTimezone('utc')
+ dashboard.withRefresh('30s')
+ dashboard.withUid(std.md5(uid))
+ dashboard.graphTooltip.withSharedCrosshair()
+ dashboard.withVariables(variables)
+ dashboard.withPanels(panels),
},
}