Skip to content

:hedgehog: PostHog is the leading platform for building self-driving products. Our developer tools – AI observability, analytics, session replay, flags, experiments, error tracking, logs, and more – capture all the context agents need to diagnose problems, uncover opportunities, and ship fixes. Steer it all from Slack, web, desktop, or the MCP.

3.0 KB
Raw Blame
# Golden-chart (charts/posthog-app) values for pgcollector.
# Copy to charts/apps/pgcollector/values.yaml and adjust the psql: entries.

image:
  name: pgcollector
  repository: 795637471508.dkr.ecr.us-east-1.amazonaws.com/pgcollector

command: ["/pgcollector"]
args: ["--config", "/etc/pgcollector/pgcollector.toml"]

role: k8s-pgcollector

# Worker: no HTTP surface except probes/metrics on :9187.
service:
  enabled: false
autoscaling:
  minPods: 1
  maxPods: 1
monitoring:
  prometheus:
    path: /metrics
    port: "9187"
livenessProbe:
  httpGet: { path: /healthz, port: 9187 }
readinessProbe:
  httpGet: { path: /readyz, port: 9187 }

resources:
  requests: { cpu: 250m, memory: 256Mi }
  limits:   { cpu: 250m, memory: 512Mi }

# One psql: entry for the sink, then one per monitored cluster.
#
# pgbouncer: false is REQUIRED on every entry — pgcollector sets per-session
# statement_timeout/lock_timeout and refuses to start behind a transaction pooler.
#
# Monitored clusters need no [[servers]] config at all: the env var names below are
# the convention pgcollector discovers servers from:
#   PGCOLLECTOR_SERVER_<ID>_URL                    writer / cluster endpoint
#   PGCOLLECTOR_SERVER_<ID>_READER_URL             load-balanced reader endpoint
#   PGCOLLECTOR_SERVER_<ID>_INSTANCE_<NAME>_URL    per-instance endpoint (optional)
#   PGCOLLECTOR_SERVER_<ID>_LOG_GROUP              CloudWatch log group for the logs collector
# <ID> and <NAME> are lower-cased with _ → - to form server_id / instance.
psql:
  pgcollector:                   # the sink; user must be in the `ddl` tier
    pgbouncer: false
    write:
      enabled: true
      env:
        PGCOLLECTOR_SINK_URL: POSTGRES_URL

  cloud:                         # monitored cluster; user in the `monitor` tier
    pgbouncer: false
    write:
      enabled: true
      env:
        PGCOLLECTOR_SERVER_CLOUD_URL: POSTGRES_URL
    read:
      enabled: true
      env:
        PGCOLLECTOR_SERVER_CLOUD_READER_URL: POSTGRES_URL

  persons:
    pgbouncer: false
    write:
      enabled: true
      env:
        PGCOLLECTOR_SERVER_PERSONS_URL: POSTGRES_URL
    read:
      enabled: true
      env:
        PGCOLLECTOR_SERVER_PERSONS_READER_URL: POSTGRES_URL

env:
  PGCOLLECTOR_SERVER_CLOUD_LOG_GROUP: /aws/rds/cluster/<cloud-cluster-id>/postgresql
  PGCOLLECTOR_SERVER_PERSONS_LOG_GROUP: /aws/rds/cluster/<persons-cluster-id>/postgresql

configFiles:
  pgcollector:
    path: /etc/pgcollector/pgcollector.toml
    content: |
      [sink]
      database_url = "${PGCOLLECTOR_SINK_URL}"
      retention_days = 14
      [sink.retention]
      query_durations = 7        # sampled statement log; the biggest table by far

      [defaults]
      databases = ["*"]            # discover every database on each cluster, keep discovering
      rediscover_interval = "10m"

      # Global per-collector tuning; [[servers]] blocks (optional) can override per cluster.
      [defaults.overrides]
      # sizes = { interval = "30m" }
      # settings = { enabled = false }