prometheus/documentation/prometheus-mixin/alerts.libsonnet

193 lines
6.9 KiB
Plaintext
Raw Normal View History

{
prometheusAlerts+:: {
groups+: [
{
name: 'prometheus',
rules: [
{
alert: 'PrometheusBadConfig',
expr: |||
prometheus_config_last_reload_successful{%(prometheusSelector)s} == 0
||| % $._config,
'for': '15m',
labels: {
severity: 'critical',
},
annotations: {
message: 'Prometheus failed to reload config, see container logs',
},
},
{
alert: 'PrometheusNotificationQueueRunningFull',
expr: |||
predict_linear(prometheus_notifications_queue_length{%(prometheusSelector)s}[5m], 60 * 30)
>
prometheus_notifications_queue_capacity{%(prometheusSelector)s}
||| % $._config,
'for': '15m',
labels: {
severity: 'warning',
},
annotations: {
message: "Prometheus' alert notification queue is running full for {{$labels.namespace}}/{{ $labels.pod}}",
},
},
{
alert: 'PrometheusErrorSendingAlerts',
expr: |||
100 * rate(prometheus_notifications_errors_total{%(prometheusSelector)s}[5m])
/
rate(prometheus_notifications_sent_total{%(prometheusSelector)s}[5m]) > 1
||| % $._config,
'for': '15m',
labels: {
severity: 'warning',
},
annotations: {
message: '{{ printf "%.1f" $value }}% errors while sending alerts from Prometheus {{$labels.namespace}}/{{ $labels.pod}} to Alertmanager {{$labels.Alertmanager}}',
},
},
{
alert: 'PrometheusErrorSendingAlerts',
expr: |||
100 * rate(prometheus_notifications_errors_total{%(prometheusSelector)s}[5m])
/
rate(prometheus_notifications_sent_total{%(prometheusSelector)s}[5m]) > 3
||| % $._config,
'for': '15m',
labels: {
severity: 'critical',
},
annotations: {
message: '{{ printf "%.1f" $value }}% errors while sending alerts from Prometheus {{$labels.namespace}}/{{ $labels.pod}} to Alertmanager {{$labels.Alertmanager}}',
},
},
{
alert: 'PrometheusNotConnectedToAlertmanagers',
expr: |||
prometheus_notifications_alertmanagers_discovered{%(prometheusSelector)s} < 1
||| % $._config,
'for': '10m',
labels: {
severity: 'warning',
},
annotations: {
message: 'Prometheus {{ $labels.namespace }}/{{ $labels.pod}} is not connected to any Alertmanagers',
},
},
{
alert: 'PrometheusTSDBReloadsFailing',
expr: |||
increase(prometheus_tsdb_reloads_failures_total{%(prometheusSelector)s}[2h]) > 0
||| % $._config,
'for': '12h',
labels: {
severity: 'warning',
},
annotations: {
message: '{{$labels.job}} at {{$labels.instance}} had {{$value | humanize}} reload failures over the last four hours.',
},
},
{
alert: 'PrometheusTSDBCompactionsFailing',
expr: |||
increase(prometheus_tsdb_compactions_failed_total{%(prometheusSelector)s}[2h]) > 0
||| % $._config,
'for': '12h',
labels: {
severity: 'warning',
},
annotations: {
message: '{{$labels.job}} at {{$labels.instance}} had {{$value | humanize}} compaction failures over the last four hours.',
},
},
{
alert: 'PrometheusTSDBWALCorruptions',
expr: |||
tsdb_wal_corruptions_total{%(prometheusSelector)s} > 0
||| % $._config,
'for': '4h',
labels: {
severity: 'warning',
},
annotations: {
message: '{{$labels.job}} at {{$labels.instance}} has a corrupted write-ahead log (WAL).',
},
},
{
alert: 'PrometheusNotIngestingSamples',
expr: |||
rate(prometheus_tsdb_head_samples_appended_total{%(prometheusSelector)s}[5m]) <= 0
||| % $._config,
'for': '10m',
labels: {
severity: 'warning',
},
annotations: {
message: "Prometheus {{ $labels.namespace }}/{{ $labels.pod}} isn't ingesting samples.",
},
},
{
alert: 'PrometheusTargetScrapesDuplicate',
expr: |||
increase(prometheus_target_scrapes_sample_duplicate_timestamp_total{%(prometheusSelector)s}[5m]) > 0
||| % $._config,
'for': '10m',
labels: {
severity: 'warning',
},
annotations: {
message: '{{$labels.namespace}}/{{$labels.pod}} has many samples rejected due to duplicate timestamps but different values',
},
},
{
alert: 'PrometheusRemoteStorageFailures',
expr: |||
(rate(prometheus_remote_storage_failed_samples_total{%(prometheusSelector)s}[1m]) * 100)
/
(rate(prometheus_remote_storage_failed_samples_total{%(prometheusSelector)s}[1m]) + rate(prometheus_remote_storage_succeeded_samples_total{%(prometheusSelector)s}[1m]))
> 1
||| % $._config,
'for': '15m',
labels: {
severity: 'critical',
},
annotations: {
message: 'Prometheus failed to send {{ printf "%.1f" $value }}% samples',
},
},
{
alert: 'PrometheusRemoteWriteBehind',
expr: |||
prometheus_remote_storage_highest_timestamp_in_seconds{%(prometheusSelector)s}
- on(job, instance) group_right
prometheus_remote_storage_queue_highest_sent_timestamp_seconds{%(prometheusSelector)s}
> 60
||| % $._config,
'for': '15m',
labels: {
severity: 'critical',
},
annotations: {
message: 'Prometheus remote write is {{ printf "%.1f" $value }}s behind.',
},
},
{
alert: 'PrometheusRuleFailures',
'for': '15m',
expr: |||
rate(prometheus_rule_evaluation_failures_total{%(prometheusSelector)s}[1m]) > 0
||| % $._config,
labels: {
severity: 'critical',
},
annotations: {
message: 'Prometheus failed to evaluate {{ printf "%.1f" $value }} rules / s',
},
},
],
},
],
},
}