跳到主要内容

Whizard 参数配置

注: 自 Whizard v0.10 起,Whizard 主要参数配置在 CRD services.monitoring.whizard.io 中。

Whizard 配置参数存储于 CRD 中,其中有组件自身的 CRD,如 queries.monitoring.whizard.ioingesters.monitoring.whizard.io, 也有定义整个 Whizard 各服务部署的 CRD services.monitoring.whizard.io, 由于 Whizard 部分组件根据租户数量进行自动伸缩, 在组件 CRD 中配置无法适应自定伸缩场景,所以在 services.monitoring.whizard.io CRD 中定义了各组件的配置模板 xxxxTemplateSpec,可以在该 CRD 中定义各组件的配置参数,但组件 CRD 配置优先级高于 xxxxTemplateSpec

spec:
compactorTemplateSpec:
dataVolume:
persistentVolumeClaim:
spec:
resources:
requests:
storage: 20Gi
# 在配置对象存储情况下每个 Compactor 实例允许处理 Tenant 数量,即 Compactor 承载租户数据的能力,默认值为10
defaultTenantsPerCompactor: 10
flags:
- --block-files-concurrency=20
- --compact.blocks-fetch-concurrency=5
- --web.disable
- --compact.enable-vertical-compaction
image: docker.io/thanosio/thanos:v0.36.0
retention:
retention1h: 360d
retention5m: 180d
retentionRaw: 60d
securityContext:
fsGroup: 0
defaultTenantId: unknown
gatewayTemplateSpec:
image: docker.io/kubesphere/whizard-monitoring-gateway:v0.10.0
ingesterTemplateSpec:
dataVolume:
persistentVolumeClaim:
spec:
resources:
requests:
storage: 20Gi
# 空租户时 Ingester 回收周期; Ingester 会由于租户删除或迁移,导致 Ingester 上没有租户,此时需要回收;
# 在使用对象存储时默认为3h, 在无对象存储情况下,此参数会和该 Ingester 的 localTsdbRetention 保持一致。
defaultIngesterRetentionPeriod: 3h
# 每个 Ingester 实例允许处理 Tenant 数量,即 Ingester 承载租户数据的能力,默认值为3
defaultTenantsPerIngester: 3
disableTsdbCleanup: true
flags:
- --tsdb.out-of-order.time-window=10m
image: docker.io/thanosio/thanos:v0.36.0
securityContext:
fsGroup: 0
queryFrontendTemplateSpec:
flags:
- --no-query-range.align-range-with-step
image: docker.io/thanosio/thanos:v0.36.0
queryTemplateSpec:
envoy:
image: docker.io/envoyproxy/envoy:v1.20.2
flags:
- --query.max-concurrent=200
image: docker.io/thanosio/thanos:v0.36.0
replicaLabelNames:
- prometheus_replica
- receive_replica
- ruler_replica
routerTemplateSpec:
image: docker.io/thanosio/thanos:v0.36.0
rulerTemplateSpec:
dataVolume:
emptyDir: {}
disableAlertingRulesAutoSelection: true
evaluationInterval: 1m
image: docker.io/thanosio/thanos:v0.36.0
prometheusConfigReloader:
image: quay.io/prometheus-operator/prometheus-config-reloader:v0.75.1
ruleSelectors:
- matchLabels:
role: alert-rules
rulerQueryProxy:
image: docker.io/kubesphere/whizard-monitoring-gateway:v0.10.0
rulerWriteProxy:
image: docker.io/kubesphere/cortex-tenant:v1.12.5
securityContext:
fsGroup: 0
shards: 1
storeTemplateSpec:
dataVolume:
persistentVolumeClaim:
spec:
resources:
requests:
storage: 20Gi
flags:
- --web.disable
- --no-cache-index-header
image: docker.io/thanosio/thanos:v0.36.0
securityContext:
fsGroup: 0
timeRanges:
- maxTime: -36h
tenantHeader: cluster
tenantLabelName: cluster