openstack-helm-infra/prometheus/values_overrides/nodes.yaml

244 lines
12 KiB
YAML

conf:
prometheus:
rules:
nodes:
groups:
- name: node.recording_rules
rules:
- record: node_filesystem_free_percent
expr: 100 * {fstype =~ "xfs|ext[34]"} / node_filesystem_size{fstype =~ "xfs|ext[34]"}
- record: node_ram_usage_percent
expr: 100 * (node_memory_MemFree + node_memory_Buffers + node_memory_Cached) / node_memory_MemTotal
- record: node_swap_usage_percent
expr: 100 * (node_memory_SwapFree + node_memory_SwapCached) / node_memory_SwapTotal
- name: nodes.alerting_rules
rules:
- alert: prom_exporter_node_unavailable
expr: avg_over_time(up{job="node-exporter"}[5m]) == 0
for: 5m
labels:
severity: warning
annotations:
description: node exporter is not collecting metrics or is not available for past 10 minutes
title: node exporter is not collecting metrics or is not available
- alert: node_filesystem_full_80percent
expr: avg_over_time(node_filesystem_free_percent[2m]) > 80
for: 5m
labels:
severity: page
annotations:
description: '{{$labels.alias}} device {{$labels.device}} on {{$labels.mountpoint}}
has less than 20% free space left.'
summary: '{{$labels.alias}}: Filesystem is running out of space soon.'
- alert: node_filesystem_full_in_4h
expr: predict_linear(node_filesystem_free{fstype =~ "xfs|ext[34]"}[1h], 4 * 3600) <= 0
for: 5m
labels:
severity: page
annotations:
description: '{{$labels.alias}} device {{$labels.device}} on {{$labels.mountpoint}}
is running out of space of in approx. 4 hours'
summary: '{{$labels.alias}}: Filesystem is running out of space in 4 hours.'
- alert: node_filedescriptors_full_in_3h
expr: predict_linear(node_filefd_allocated[1h], 3 * 3600) >= node_filefd_maximum
for: 20m
labels:
severity: page
annotations:
description: '{{$labels.alias}} is running out of available file descriptors
in approx. 3 hours'
summary: '{{$labels.alias}} is running out of available file descriptors in
3 hours.'
- alert: node_load1_90percent
expr: node_load1 / ON(alias) count(node_cpu{mode="system"}) BY (alias) >= 0.9
for: 1h
labels:
severity: page
annotations:
description: '{{$labels.alias}} is running with > 90% total load for at least
1h.'
summary: '{{$labels.alias}}: Running on high load.'
- alert: node_cpu_util_90percent
expr: 100 - (avg(irate(node_cpu{mode="idle"}[5m])) BY (alias) * 100) >= 90
for: 1h
labels:
severity: page
annotations:
description: '{{$labels.alias}} has total CPU utilization over 90% for at least
1h.'
summary: '{{$labels.alias}}: High CPU utilization.'
- alert: node_ram_using_90percent
expr: avg_over_time(node_ram_usage_percent[2m]) > 90
for: 30m
labels:
severity: page
annotations:
description: '{{$labels.alias}} is using at least 90% of its RAM for at least
30 minutes now.'
summary: '{{$labels.alias}}: Using lots of RAM.'
- alert: node_swap_using_80percent
expr: avg_over_time(node_swap_usage_percent[2m]) > 80
for: 10m
labels:
severity: page
annotations:
description: '{{$labels.alias}} is using 80% of its swap space for at least
10 minutes now.'
summary: '{{$labels.alias}}: Running out of swap soon.'
- alert: node_high_cpu_load
expr: node_load15 / on(alias) count(node_cpu{mode="system"}) by (alias) >= 0
for: 1m
labels:
severity: warning
annotations:
description: '{{$labels.alias}} is running with load15 > 1 for at least 5 minutes: {{$value}}'
summary: '{{$labels.alias}}: Running on high load: {{$value}}'
- alert: node_high_memory_load
expr: avg_over_time(node_ram_usage_percent[2m]) > 85
for: 1m
labels:
severity: warning
annotations:
description: Host memory usage is {{ humanize $value }}%. Reported by
instance {{ $labels.instance }} of job {{ $labels.job }}.
summary: Server memory is almost full
- alert: node_high_storage_load
expr: avg_over_time(node_storage_usage_percent{mountpoint="/"}[2m]) > 85
for: 30s
labels:
severity: warning
annotations:
description: Host storage usage is {{ humanize $value }}%. Reported by
instance {{ $labels.instance }} of job {{ $labels.job }}.
summary: Server storage is almost full
- alert: node_high_swap
expr: (node_memory_SwapTotal - node_memory_SwapFree) < (node_memory_SwapTotal
* 0.4)
for: 1m
labels:
severity: warning
annotations:
description: Host system has a high swap usage of {{ humanize $value }}. Reported
by instance {{ $labels.instance }} of job {{ $labels.job }}.
summary: Server has a high swap usage
- alert: node_high_network_drop_rcv
expr: node_network_receive_drop{device!="lo"} > 3000
for: 30s
labels:
severity: warning
annotations:
description: Host system has an unusally high drop in network reception ({{
humanize $value }}). Reported by instance {{ $labels.instance }} of job {{
$labels.job }}
summary: Server has a high receive drop
- alert: node_high_network_drop_send
expr: node_network_transmit_drop{device!="lo"} > 3000
for: 30s
labels:
severity: warning
annotations:
description: Host system has an unusally high drop in network transmission ({{
humanize $value }}). Reported by instance {{ $labels.instance }} of job {{
$labels.job }}
summary: Server has a high transmit drop
- alert: node_high_network_errs_rcv
expr: node_network_receive_errs{device!="lo"} > 3000
for: 30s
labels:
severity: warning
annotations:
description: Host system has an unusally high error rate in network reception
({{ humanize $value }}). Reported by instance {{ $labels.instance }} of job
{{ $labels.job }}
summary: Server has unusual high reception errors
- alert: node_high_network_errs_send
expr: node_network_transmit_errs{device!="lo"} > 3000
for: 30s
labels:
severity: warning
annotations:
description: Host system has an unusally high error rate in network transmission
({{ humanize $value }}). Reported by instance {{ $labels.instance }} of job
{{ $labels.job }}
summary: Server has unusual high transmission errors
- alert: node_network_conntrack_usage_80percent
expr: sort(node_nf_conntrack_entries{job="node-exporter"} > node_nf_conntrack_entries_limit{job="node-exporter"} * 0.8)
for: 5m
labels:
severity: page
annotations:
description: '{{$labels.instance}} has network conntrack entries of {{ $value }} which is more than 80% of maximum limit'
summary: '{{$labels.instance}}: available network conntrack entries are low.'
- alert: node_entropy_available_low
expr: node_entropy_available_bits < 300
for: 5m
labels:
severity: page
annotations:
description: '{{$labels.instance}} has available entropy bits of {{ $value }} which is less than required of 300'
summary: '{{$labels.instance}}: is low on entropy bits.'
- alert: node_hwmon_high_cpu_temp
expr: node_hwmon_temp_crit_celsius*0.9 - node_hwmon_temp_celsius < 0 OR node_hwmon_temp_max_celsius*0.95 - node_hwmon_temp_celsius < 0
for: 5m
labels:
severity: page
annotations:
description: '{{$labels.alias}} reports hwmon sensor {{$labels.sensor}}/{{$labels.chip}} temperature value is nearly critical: {{$value}}'
summary: '{{$labels.alias}}: Sensor {{$labels.sensor}}/{{$labels.chip}} temp is high: {{$value}}'
- alert: node_vmstat_paging_rate_high
expr: irate(node_vmstat_pgpgin[5m]) > 80
for: 5m
labels:
severity: page
annotations:
description: '{{$labels.alias}} has a memory paging rate of change higher than 80%: {{$value}}'
summary: '{{$labels.alias}}: memory paging rate is high: {{$value}}'
- alert: node_xfs_block_allocation_high
expr: 100*(node_xfs_extent_allocation_blocks_allocated_total{job="node-exporter", instance=~"172.17.0.1.*"} / (node_xfs_extent_allocation_blocks_freed_total{job="node-exporter", instance=~"172.17.0.1.*"} + node_xfs_extent_allocation_blocks_allocated_total{job="node-exporter", instance=~"172.17.0.1.*"})) > 80
for: 5m
labels:
severity: page
annotations:
description: '{{$labels.alias}} has xfs allocation blocks higher than 80%: {{$value}}'
summary: '{{$labels.alias}}: xfs block allocation high: {{$value}}'
- alert: node_network_bond_slaves_down
expr: node_net_bonding_slaves - node_net_bonding_slaves_active > 0
for: 5m
labels:
severity: page
annotations:
description: '{{ $labels.master }} is missing {{ $value }} slave interface(s).'
summary: 'Instance {{ $labels.instance }}: {{ $labels.master }} missing {{ $value }} slave interface(s)'
- alert: node_numa_memory_used
expr: 100*node_memory_numa_MemUsed / node_memory_numa_MemTotal > 80
for: 5m
labels:
severity: page
annotations:
description: '{{$labels.alias}} has more than 80% NUMA memory usage: {{ $value }}'
summary: '{{$labels.alias}}: has high NUMA memory usage: {{$value}}'
- alert: node_ntp_clock_skew_high
expr: abs(node_ntp_drift_seconds) > 2
for: 5m
labels:
severity: page
annotations:
description: '{{$labels.alias}} has time difference of more than 2 seconds compared to NTP server: {{ $value }}'
summary: '{{$labels.alias}}: time is skewed by : {{$value}} seconds'
- alert: node_disk_read_latency
expr: (rate(node_disk_read_time_ms[5m]) / rate(node_disk_reads_completed[5m])) > 40
for: 5m
labels:
severity: page
annotations:
description: '{{$labels.device}} has a high read latency of {{ $value }}'
summary: 'High read latency observed for device {{ $labels.device }}'
- alert: node_disk_write_latency
expr: (rate(node_disk_write_time_ms[5m]) / rate(node_disk_writes_completed[5m])) > 40
for: 5m
labels:
severity: page
annotations:
description: '{{$labels.device}} has a high write latency of {{ $value }}'
summary: 'High write latency observed for device {{ $labels.device }}'