search

grafana / loki

Like Prometheus, but for logs.
https://grafana.com/loki
GNU Affero General Public License v3.0
23.28k stars 3.37k forks source link

Compactor retention with tsdb_shipper does not work #11811

Open pyo-counting opened 7 months ago

pyo-counting commented 7 months ago

Describe the bug compactor retention does not work with tsdb shipper, AWS s3 object storage.

To Reproduce Steps to reproduce the behavior:

  1. Install Helm chart with custom value file(ssd-values.yaml)

    helm install loki grafana/loki --version 5.41.8 --namespace loki-ns -f ssd-values.yaml
    # ssd-values.yaml
loki:
  # -- The number of old ReplicaSets to retain to allow rollback
  revisionHistoryLimit: 2
  # -- Config file contents for Loki
  # @default -- See values.yaml
  config: |
    {{- if .Values.enterprise.enabled}}
    {{- tpl .Values.enterprise.config . }}
    {{- else }}
    auth_enabled: {{ .Values.loki.auth_enabled }}
    {{- end }}

    {{- with .Values.loki.server }}
    server:
      {{- toYaml . | nindent 2}}
    {{- end}}

    memberlist:
    {{- if .Values.loki.memberlistConfig }}
      {{- toYaml .Values.loki.memberlistConfig | nindent 2 }}
    {{- else }}
    {{- if .Values.loki.extraMemberlistConfig}}
    {{- toYaml .Values.loki.extraMemberlistConfig | nindent 2}}
    {{- end }}
      join_members:
        - {{ include "loki.memberlist" . }}
        {{- with .Values.migrate.fromDistributed }}
        {{- if .enabled }}
        - {{ .memberlistService }}
        {{- end }}
        {{- end }}
    {{- end }}

    {{- with .Values.loki.ingester }}
    ingester:
      {{- tpl (. | toYaml) $ | nindent 4 }}
    {{- end }}

    {{- if .Values.loki.commonConfig}}
    common:
    {{- toYaml .Values.loki.commonConfig | nindent 2}}
      storage:
      {{- include "loki.commonStorageConfig" . | nindent 4}}
    {{- end}}

    {{- with .Values.loki.limits_config }}
    limits_config:
      {{- tpl (. | toYaml) $ | nindent 4 }}
    {{- end }}

    runtime_config:
      file: /etc/loki/runtime-config/runtime-config.yaml

    {{- with .Values.loki.memcached.chunk_cache }}
    {{- if and .enabled (or .host .addresses) }}
    chunk_store_config:
      chunk_cache_config:
        memcached:
          batch_size: {{ .batch_size }}
          parallelism: {{ .parallelism }}
        memcached_client:
          {{- if .host }}
          host: {{ .host }}
          {{- end }}
          {{- if .addresses }}
          addresses: {{ .addresses }}
          {{- end }}
          service: {{ .service }}
    {{- end }}
    {{- end }}

    {{- if .Values.loki.schemaConfig }}
    schema_config:
    {{- toYaml .Values.loki.schemaConfig | nindent 2}}
    {{- else }}
    schema_config:
      configs:
        - from: 2022-01-11
          store: boltdb-shipper
          object_store: {{ .Values.loki.storage.type }}
          schema: v12
          index:
            prefix: loki_index_
            period: 24h
    {{- end }}

    {{ include "loki.rulerConfig" . }}

    {{- if or .Values.tableManager.retention_deletes_enabled .Values.tableManager.retention_period }}
    table_manager:
      retention_deletes_enabled: {{ .Values.tableManager.retention_deletes_enabled }}
      retention_period: {{ .Values.tableManager.retention_period }}
    {{- end }}

    {{- with .Values.loki.memcached.results_cache }}
    query_range:
      align_queries_with_step: true
      {{- if and .enabled (or .host .addresses) }}
      cache_results: {{ .enabled }}
      results_cache:
        cache:
          default_validity: {{ .default_validity }}
          memcached_client:
            {{- if .host }}
            host: {{ .host }}
            {{- end }}
            {{- if .addresses }}
            addresses: {{ .addresses }}
            {{- end }}
            service: {{ .service }}
            timeout: {{ .timeout }}
      {{- end }}
    {{- end }}

    {{- with .Values.loki.storage_config }}
    storage_config:
      {{- tpl (. | toYaml) $ | nindent 4 }}
    {{- end }}

    {{- with .Values.loki.query_scheduler }}
    query_scheduler:
      {{- tpl (. | toYaml) $ | nindent 4 }}
    {{- end }}

    {{- with .Values.loki.compactor }}
    compactor:
      {{- tpl (. | toYaml) $ | nindent 4 }}
    {{- end }}

    {{- with .Values.loki.analytics }}
    analytics:
      {{- tpl (. | toYaml) $ | nindent 4 }}
    {{- end }}

    {{- with .Values.loki.querier }}
    querier:
      {{- tpl (. | toYaml) $ | nindent 4 }}
    {{- end }}

    {{- with .Values.loki.index_gateway }}
    index_gateway:
      {{- tpl (. | toYaml) $ | nindent 4 }}
    {{- end }}

    {{- with .Values.loki.frontend }}
    frontend:
      {{- tpl (. | toYaml) $ | nindent 4 }}
    {{- end }}

    {{- with .Values.loki.frontend_worker }}
    frontend_worker:
      {{- tpl (. | toYaml) $ | nindent 4 }}
    {{- end }}

    {{- with .Values.loki.distributor }}
    distributor:
      {{- tpl (. | toYaml) $ | nindent 4 }}
    {{- end }}

    tracing:
      enabled: {{ .Values.loki.tracing.enabled }}
  # Should authentication be enabled
  auth_enabled: true
  # -- Check https://grafana.com/docs/loki/latest/configuration/#server for more info on the server configuration.
  server:
    log_format: "logfmt"
    log_level: "info"
    log_source_ips_enabled: true
    log_request_headers: true
    log_request_at_info_level_enabled: true
  # -- Limits config
  limits_config:
    max_line_size: 10KB
    per_stream_rate_limit: 5MB
    per_stream_rate_limit_burst: 20MB
    split_queries_by_interval: 15m
    retention_period: 7d
    retention_stream:
      - selector: '{environment="dev"}'
        priority: 1
        period: 1d
      - selector: '{environment="stg"}'
        priority: 1
        period: 2d
    shard_streams:
      enabled: false
    allow_structured_metadata: true
  # -- Provides a reloadable runtime configuration file for some specific configuration
  runtimeConfig: {}
  # -- Check https://grafana.com/docs/loki/latest/configuration/#common_config for more info on how to provide a common     configuration
  commonConfig:
    path_prefix: /var/loki
    replication_factor: 3
    ring:
      kvstore:
        store: "memberlist"
    compactor_address: '{{ include "loki.compactorAddress" . }}'
  # -- Storage config. Providing this will automatically populate all necessary storage configs in the templated config.
  storage:
    bucketNames:
      chunks: kps-shr-tools-s3-loki
      ruler: kps-shr-tools-s3-loki
    type: s3
    s3:
      region: ap-northeast-2
  # -- Configure memcached as an external cache for chunk and results cache. Disabled by default
  # must enable and specify a host for each cache you would like to use.
  memcached:
    chunk_cache:
      enabled: false
    results_cache:
      enabled: false
  # -- Check https://grafana.com/docs/loki/latest/configuration/#schema_config for more info on how to configure schemas
  schemaConfig:
    configs:
      - from: "2024-01-01"
        store: tsdb
        object_store: s3
        schema: v13
        index:
          prefix: tsdb_index_
          period: 24h
  # -- Check https://grafana.com/docs/loki/latest/configuration/#ruler for more info on configuring ruler
  rulerConfig: {}
  # -- Structured loki configuration, takes precedence over `loki.config`, `loki.schemaConfig`, `loki.storageConfig`
  structuredConfig:
    common:
      storage:
        s3:
          storage_class: "STANDARD"
        hedging:
          at: 250ms
          up_to: 3
          max_per_second: 20
    query_range:
      results_cache:
        cache:
          enable_fifocache: false
          embedded_cache:
            enabled: true
            max_size_mb: 150
            ttl: 30m
        compression: "snappy"
      cache_results: true
      cache_index_stats_results: false
  # -- Additional query scheduler config
  query_scheduler:
    max_outstanding_requests_per_tenant: 32768
    querier_forget_delay: 60s
  # -- Additional storage config
  storage_config:
    aws:
      bucketnames: kps-shr-tools-s3-loki
      region: ap-northeast-2
      insecure: false
      storage_class: "STANDARD"
    tsdb_shipper:
      active_index_directory: /var/loki/ingester/tsdb_shipper
      shared_store: "s3"
      shared_store_key_prefix: "tsdb_shipper/"
      cache_location: /var/loki/index_gateway/tsdb_shipper
      index_gateway_client:
        log_gateway_requests: true
  # --  Optional compactor configuration
  compactor:
    working_directory: "/var/loki/compactor"
    shared_store: "s3"
    shared_store_key_prefix: "compoactor/"
    retention_enabled: true
    compactor_ring:
      kvstore:
        store: "memberlist"
  # --  Optional analytics configuration
  analytics:
    reporting_enabled: false
  # --  Optional querier configuration
  querier:
    tail_max_duration: 30m
    max_concurrent: 16
    multi_tenant_queries_enabled: true
  # --  Optional ingester configuration
  ingester:
    lifecycler:
      ring:
        kvstore:
          store: "memberlist"
      final_sleep: 15s
    wal:
      dir: "/var/loki/ingester/wal"
      flush_on_shutdown: true
      replay_memory_ceiling: 1GB
  # --  Optional index gateway configuration
  index_gateway:
    mode: ring
    ring:
      kvstore:
        store: "memberlist"
  frontend:
    scheduler_address: '{{ include "loki.querySchedulerAddress" . }}'
    log_queries_longer_than: 5s
    query_stats_enabled: true
    scheduler_dns_lookup_period: 3s
    compress_responses: true
  frontend_worker:
    match_max_concurrent: true
    scheduler_address: '{{ include "loki.querySchedulerAddress" . }}'
  # -- Optional distributor configuration
  distributor:
    ring:
      kvstore:
        store: "memberlist"
    rate_store:
      debug: true
    write_failures_logging:
      add_insights_label: true
  # -- Enable tracing
  tracing:
    enabled: true
enterprise:
  # Enable enterprise features, license must be provided
  enabled: false

# -- Options that may be necessary when performing a migration from another helm chart
migrate:
  # -- When migrating from a distributed chart like loki-distributed or enterprise-logs
  fromDistributed:
    # -- Set to true if migrating from a distributed helm chart
    enabled: false

serviceAccount:
  # -- Specifies whether a ServiceAccount should be created
  create: true
  # -- The name of the ServiceAccount to use.
  # If not set and create is true, a name is generated using the fullname template
  name: loki-sa
  # -- Annotations for the service account
  annotations:
    eks.amazonaws.com/role-arn: (...skip...)
  # -- Set this toggle to false to opt out of automounting API credentials for the service account
  automountServiceAccountToken: true

# RBAC configuration
rbac:
  # -- If pspEnabled true, a PodSecurityPolicy is created for K8s that use psp.
  pspEnabled: false
  # -- For OpenShift set pspEnabled to 'false' and sccEnabled to 'true' to use the SecurityContextConstraints.
  sccEnabled: false

# -- Section for configuring optional Helm test
test:
  enabled: false

# Monitoring section determines which monitoring features to enable
monitoring:
  # Dashboards for monitoring Loki
  dashboards:
    # -- If enabled, create configmap with dashboards for monitoring Loki
    enabled: false
  # Recording rules for monitoring Loki, required for some dashboards
  rules:
    # -- If enabled, create PrometheusRule resource with Loki recording rules
    enabled: false
    # -- Include alerting rules
    alerting: false
  # ServiceMonitor configuration
  serviceMonitor:
    # -- If enabled, ServiceMonitor resources for Prometheus Operator are created
    enabled: false
  # Self monitoring determines whether Loki should scrape its own logs.
  # This feature currently relies on the Grafana Agent Operator being installed,
  # which is installed by default using the grafana-agent-operator sub-chart.
  # It will create custom resources for GrafanaAgent, LogsInstance, and PodLogs to configure
  # scrape configs to scrape its own logs with the labels expected by the included dashboards.
  selfMonitoring:
    enabled: false
  # The Loki canary pushes logs to and queries from this loki installation to test
  # that it's working correctly
  lokiCanary:
    enabled: false

# Configuration for the write pod(s)
write:
  # -- Number of replicas for the write
  replicas: 3
  autoscaling:
    # -- Enable autoscaling for the write.
    enabled: false
  # -- Comma-separated list of Loki modules to load for the write
  targetModule: "write"
  # -- Resource requests and limits for the write
  resources:
    limits:
      cpu: 1.5
      memory: 2Gi
    requests:
      cpu: 500m
      memory: 500Mi
  # -- Grace period to allow the write to shutdown before it is killed. Especially for the ingester,
  # this must be increased. It must be long enough so writes can be gracefully shutdown flushing/transferring
  # all data and to successfully leave the member ring on shutdown.
  terminationGracePeriodSeconds: 300
  # -- The default is to deploy all pods in parallel.
  podManagementPolicy: "Parallel"
  persistence:
    # -- Enable volume claims in pod spec
    volumeClaimsEnabled: true
    # -- Enable StatefulSetAutoDeletePVC feature
    enableStatefulSetAutoDeletePVC: false
    # -- Storage class to be used.
    # If defined, storageClassName: <storageClass>.
    # If set to "-", storageClassName: "", which disables dynamic provisioning.
    # If empty or set to null, no storageClassName spec is
    # set, choosing the default provisioner (gp2 on AWS, standard on GKE, AWS, and OpenStack).
    storageClass: loki-sc

# Configuration for the table-manager
tableManager:
  # -- Specifies whether the table-manager should be enabled
  enabled: false

# Configuration for the read pod(s)
read:
  # -- Number of replicas for the read
  replicas: 2
  autoscaling:
    # -- Enable autoscaling for the read, this is only used if `queryIndex.enabled: true`
    enabled: false
  # -- Comma-separated list of Loki modules to load for the read
  targetModule: "read"
  # -- Whether or not to use the 2 target type simple scalable mode (read, write) or the
  # 3 target type (read, write, backend). Legacy refers to the 2 target type, so true will
  # run two targets, false will run 3 targets.
  legacyReadTarget: false
  # -- Resource requests and limits for the read
  resources:
    limits:
      cpu: 1.5
      memory: 2Gi
    requests:
      cpu: 500m
      memory: 500Mi
  # -- Grace period to allow the read to shutdown before it is killed
  terminationGracePeriodSeconds: 30

# Configuration for the backend pod(s)
backend:
  # -- Number of replicas for the backend
  replicas: 2
  autoscaling:
    # -- Enable autoscaling for the backend.
    enabled: false
  # -- Comma-separated list of Loki modules to load for the read
  targetModule: "backend"
  # -- Resource requests and limits for the backend
  resources:
    limits:
      cpu: 1
      memory: 1Gi
    requests:
      cpu: 500m
      memory: 500Mi
  # -- Grace period to allow the backend to shutdown before it is killed. Especially for the ingester,
  # this must be increased. It must be long enough so backends can be gracefully shutdown flushing/transferring
  # all data and to successfully leave the member ring on shutdown.
  terminationGracePeriodSeconds: 300
  podManagementPolicy: "Parallel"
  persistence:
    # -- Enable volume claims in pod spec
    volumeClaimsEnabled: true
    # -- Enable StatefulSetAutoDeletePVC feature
    enableStatefulSetAutoDeletePVC: true
    # -- Storage class to be used.
    # If defined, storageClassName: <storageClass>.
    # If set to "-", storageClassName: "", which disables dynamic provisioning.
    # If empty or set to null, no storageClassName spec is
    # set, choosing the default provisioner (gp2 on AWS, standard on GKE, AWS, and OpenStack).
    storageClass: loki-sc
# Configuration for the single binary node(s)
singleBinary:
  # -- Number of replicas for the single binary
  replicas: 0

# Use either this ingress or the gateway, but not both at once.
# If you enable this, make sure to disable the gateway.
# You'll need to supply authn configuration for your ingress controller.
ingress:
  enabled: true
  ingressClassName: "alb"
  annotations:
    (...skip...)
  paths:
    (...skip...)
  hosts:
    (...skip...)

# Configuration for the memberlist service
memberlist:
  service:
    publishNotReadyAddresses: false

# Configuration for the gateway
gateway:
  # -- Specifies whether the gateway should be enabled
  enabled: false

networkPolicy:
  # -- Specifies whether Network Policies should be created
  enabled: false

# -------------------------------------
# Configuration for `minio` child chart
# -------------------------------------
minio:
  enabled: false

sidecar:
  rules:
    # -- Whether or not to create a sidecar to ingest rule from specific ConfigMaps and/or Secrets.
    enabled: false
  2. push logs with Promtail to Loki(stream: {environment="dev" ...}, tenant id: kurlypay)

Expected behavior A clear and concise description of what you expected to happen.

Environment:

Screenshots, Promtail config, or terminal output If applicable, add any output to help explain your problem.

pyo-counting commented 7 months ago

I tested below values.yaml file and checked compaction and retention is working.

loki:
  auth_enabled: false
  limits_config:
    retention_period: 1d
  commonConfig:
    replication_factor: 2
  storage:
    bucketNames:
      chunks: kps-shr-tools-s3-loki-test
      ruler: kps-shr-tools-s3-loki-test
    s3:
      region: ap-northeast-2
  storage_config:
    boltdb_shipper:
        active_index_directory: /var/loki/data/index
        cache_location: /var/loki/data/boltdb-cache
        shared_store: s3
  compactor:
    working_directory: /var/loki/data/retention
    shared_store: s3
    retention_delete_delay: 30m
    compaction_interval: 10m
    retention_enabled: true
    retention_delete_worker_count: 150
serviceAccount:
  name: loki-sa
  imagePullSecrets: []
  annotations:
    eks.amazonaws.com/role-arn: (...skip...)
  rules:
    enabled: false
    alerting: false
  serviceMonitor:
    enabled: false
  lokiCanary:
    enabled: false
write:
  replicas: 2
  persistence:
    storageClass: loki-sc
read:
  replicas: 2
  persistence:
    storageClass: loki-sc
backend:
  replicas: 2
  persistence:
    storageClass: loki-sc
gateway:
  enabled: false
extraObjects:
  - apiVersion: storage.k8s.io/v1
    kind: StorageClass
    metadata:
      name: loki-sc
    provisioner: efs.csi.aws.com
    parameters:
      provisioningMode: efs-ap
      fileSystemId: (...skip...)
      directoryPerms: "700"
      uid: '{{ .Values.loki.podSecurityContext.runAsUser }}'
      gid: '{{ .Values.loki.podSecurityContext.runAsGroup }}'

What did I miss? Let me know, please

pyo-counting commented 7 months ago

Finally, I found the cause. It's a problem caused by the different value between-tsdb.shipper.shared-store.key-prefix and -compactor.shared-store.key-prefix.

I simply thought the compactor was using -compactor.shared-store.key-prefix flag for deletion purposes not for compact and retention. But it wasn't.

I hope this will be added to the official Loki documentation. There are each option for compactor and writer, so there might be people who have the same misconception as me.

icanhazbeer commented 6 months ago

Hi, Can you elaborate on this? Did you set each value separately?

pyo-counting commented 6 months ago

@icanhazbeer That's right. I set two runtime flag values to different values.

And the results of the test are as follows.

As a result, we can see that the two flags must always have the same value in order for a comacptor to perform a compaction, retention properly.