prometheus/consoles/node-overview.html
Julian Wiedmann 690d5d0b51 consoles: exclude iowait and steal from CPU Utilisation
'iowait' and 'steal' indicate specific idle/wait states, which shouldn't
be counted into CPU Utilisation. Also see
https://github.com/prometheus-operator/kube-prometheus/pull/796 and
https://github.com/kubernetes-monitoring/kubernetes-mixin/pull/667.

Per the iostat man page:

%idle
    Show the percentage of time that the CPU or CPUs were idle and the
    system did not have an outstanding disk I/O request.

%iowait
     Show the percentage of time that the CPU or CPUs were idle during
     which the system had an outstanding disk I/O request.

%steal
     Show the percentage of time spent in involuntary wait by the
     virtual CPU or CPUs while the hypervisor was servicing another
     virtual processor.

Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
2021-11-03 12:47:23 +01:00

122 lines
5.7 KiB
HTML

{{ template "head" . }}
{{ template "prom_right_table_head" }}
<tr><th colspan="2">Overview</th></tr>
<tr>
<td>User CPU</td>
<td>{{ template "prom_query_drilldown" (args (printf "sum(irate(node_cpu_seconds_total{job='node',instance='%s',mode='user'}[5m])) * 100 / count(count by (cpu)(node_cpu_seconds_total{job='node',instance='%s'}))" .Params.instance .Params.instance) "%" "printf.1f") }}</td>
</tr>
<tr>
<td>System CPU</td>
<td>{{ template "prom_query_drilldown" (args (printf "sum(irate(node_cpu_seconds_total{job='node',instance='%s',mode='system'}[5m])) * 100 / count(count by (cpu)(node_cpu_seconds_total{job='node',instance='%s'}))" .Params.instance .Params.instance) "%" "printf.1f") }}</td>
</tr>
<tr>
<td>Memory Total</td>
<td>{{ template "prom_query_drilldown" (args (printf "node_memory_MemTotal_bytes{job='node',instance='%s'}" .Params.instance) "B" "humanize1024") }}</td>
</tr>
<tr>
<td>Memory Free</td>
<td>{{ template "prom_query_drilldown" (args (printf "node_memory_MemFree_bytes{job='node',instance='%s'}" .Params.instance) "B" "humanize1024") }}</td>
</tr>
<tr>
<th colspan="2">Network</th>
</tr>
{{ range printf "node_network_receive_bytes_total{job='node',instance='%s',device!='lo'}" .Params.instance | query | sortByLabel "device" }}
<tr>
<td>{{ .Labels.device }} Received</td>
<td>{{ template "prom_query_drilldown" (args (printf "irate(node_network_receive_bytes_total{job='node',instance='%s',device='%s'}[5m])" .Labels.instance .Labels.device) "B/s" "humanize") }}</td>
</tr>
<tr>
<td>{{ .Labels.device }} Transmitted</td>
<td>{{ template "prom_query_drilldown" (args (printf "irate(node_network_transmit_bytes_total{job='node',instance='%s',device='%s'}[5m])" .Labels.instance .Labels.device) "B/s" "humanize") }}</td>
</tr>
{{ end }}
<tr>
<th colspan="2">Disks</th>
</tr>
{{ range printf "node_disk_io_time_seconds_total{job='node',instance='%s',device!~'^(md\\\\d+$|dm-)'}" .Params.instance | query | sortByLabel "device" }}
<tr>
<td>{{ .Labels.device }} Utilization</td>
<td>{{ template "prom_query_drilldown" (args (printf "irate(node_disk_io_time_seconds_total{job='node',instance='%s',device='%s'}[5m]) * 100" .Labels.instance .Labels.device) "%" "printf.1f") }}</td>
</tr>
{{ end }}
{{ range printf "node_disk_io_time_seconds_total{job='node',instance='%s'}" .Params.instance | query | sortByLabel "device" }}
<tr>
<td>{{ .Labels.device }} Throughput</td>
<td>{{ template "prom_query_drilldown" (args (printf "irate(node_disk_read_bytes_total{job='node',instance='%s',device='%s'}[5m]) + irate(node_disk_written_bytes_total{job='node',instance='%s',device='%s'}[5m])" .Labels.instance .Labels.device .Labels.instance .Labels.device) "B/s" "humanize") }}</td>
</tr>
{{ end }}
<tr>
<th colspan="2">Filesystem Fullness</th>
</tr>
{{ define "roughlyNearZero" }}
{{ if gt .1 . }}~0{{ else }}{{ printf "%.1f" . }}{{ end }}
{{ end }}
{{ range printf "node_filesystem_size_bytes{job='node',instance='%s'}" .Params.instance | query | sortByLabel "mountpoint" }}
<tr>
<td>{{ .Labels.mountpoint }}</td>
<td>{{ template "prom_query_drilldown" (args (printf "100 - node_filesystem_avail_bytes{job='node',instance='%s',mountpoint='%s'} / node_filesystem_size_bytes{job='node'} * 100" .Labels.instance .Labels.mountpoint) "%" "roughlyNearZero") }}</td>
</tr>
{{ end }}
{{ template "prom_right_table_tail" }}
{{ template "prom_content_head" . }}
<h1>Node Overview - {{ reReplaceAll "(.*?://)([^:/]+?)(:\\d+)?/.*" "$2" .Params.instance }}</h1>
<h3>CPU Usage</h3>
<div id="cpuGraph"></div>
<script>
new PromConsole.Graph({
node: document.querySelector("#cpuGraph"),
expr: "sum by (mode)(irate(node_cpu_seconds_total{job='node',instance='{{ .Params.instance }}',mode!='idle',mode!='iowait',mode!='steal'}[5m]))",
renderer: 'area',
max: {{ with printf "count(count by (cpu)(node_cpu_seconds_total{job='node',instance='%s'}))" .Params.instance | query }}{{ . | first | value }}{{ else}}undefined{{end}},
yAxisFormatter: PromConsole.NumberFormatter.humanizeNoSmallPrefix,
yHoverFormatter: PromConsole.NumberFormatter.humanizeNoSmallPrefix,
yTitle: 'Cores'
})
</script>
<h3>Disk I/O Utilization</h3>
<div id="diskioGraph"></div>
<script>
new PromConsole.Graph({
node: document.querySelector("#diskioGraph"),
expr: [
"irate(node_disk_io_time_seconds_total{job='node',instance='{{ .Params.instance }}',device!~'^(md\\\\d+$|dm-)'}[5m]) * 100",
],
min: 0,
name: '[[ device ]]',
yAxisFormatter: PromConsole.NumberFormatter.humanizeNoSmallPrefix,
yHoverFormatter: PromConsole.NumberFormatter.humanizeNoSmallPrefix,
yUnits: "%",
yTitle: 'Disk I/O Utilization'
})
</script>
<h3>Memory</h3>
<div id="memoryGraph"></div>
<script>
new PromConsole.Graph({
node: document.querySelector("#memoryGraph"),
renderer: 'area',
expr: [
"node_memory_Cached_bytes{job='node',instance='{{ .Params.instance }}'}",
"node_memory_Buffers_bytes{job='node',instance='{{ .Params.instance }}'}",
"node_memory_MemTotal_bytes{job='node',instance='{{ .Params.instance }}'} - node_memory_MemFree_bytes{job='node',instance='{{.Params.instance}}'} - node_memory_Buffers_bytes{job='node',instance='{{.Params.instance}}'} - node_memory_Cached_bytes{job='node',instance='{{.Params.instance}}'}",
"node_memory_MemFree{job='node',instance='{{ .Params.instance }}'}",
],
name: ["Cached", "Buffers", "Used", "Free"],
min: 0,
yUnits: "B",
yAxisFormatter: PromConsole.NumberFormatter.humanize1024,
yHoverFormatter: PromConsole.NumberFormatter.humanize1024,
yTitle: 'Memory'
})
</script>
{{ template "prom_content_tail" . }}
{{ template "tail" }}