Skip to content

README.md

Latest commit

 

History

History
202 lines (165 loc) · 14.9 KB

README.md

File metadata and controls

202 lines (165 loc) · 14.9 KB

Crunchy Postgres chart

A chart to provision a Crunchy Postgres cluster.

Configuration

Apply base configuration from values.yaml and make the necessary overrides in custom-values-example.yaml

helm upgrade --install new-hippo-ha . -f values.yaml -f custom-values-example.yaml

Crunchy Options

Parameter Description Default
fullnameOverride Override release name crunchy-postgres
crunchyImage Crunchy Postgres image
postgresVersion Postgres version 15

Instances

Parameter Description Default
instances.name Instance name ha (high availability)
instances.replicas Number of replicas 2
instances.dataVolumeClaimSpec.storage Amount of storage for each PVC 256Mi
instances.requests.cpu CPU requests 1m
instances.requests.memory Memory requests 256Mi
instances.limits.cpu CPU limits 100m
instances.limits.memory Memory limits 512Mi
instances.replicaCertCopy.requests.cpu replicaCertCopy CPU requests 1m
instances.replicaCertCopy.requests.memory replicaCertCopyMemory requests 32Mi
instances.replicaCertCopy.limits.cpu replicaCertCopyCPU limits 50m
instances.replicaCertCopy.limits.memory replicaCertCopy Memory limits 64Mi

pgBackRest - Reliable PostgreSQL Backup & Restore

pgBackRest site Crunchy pgBackRest docs

Parameter Description Default
pgBackRest.image Crunchy pgBackRest
pgBackRest.retention Number of backups/days to keep depending on retentionFullType 2
pgBackRest.retentionFullType Either 'count' or 'time' count
pgBackRest.repos.schedules.full Full backup schedule 0 8 * * *
pgBackRest.repos.schedules.incremental Incremental backup schedule 0 0,4,12,16,20 * * *
pgBackRest.repos.schedules.volume.addessModes Access modes ReadWriteOnce
pgBackRest.repos.schedules.volume.storage PVC size 128Mi
pgBackRest.repos.schedules.volume.storageClassName Storage class name modes netapp-file-backup
pgBackRest.repoHost.requests.cpu CPU requests 1m
pgBackRest.repoHost.requests.memory Memory requests 64Mi
pgBackRest.repoHost.limits.cpu CPU limits 50m
pgBackRest.repoHost.limits.memory Memory limits 128Mi
pgBackRest.sidecars.requests.cpu sidecars CPU requests 1m
pgBackRest.sidecars.requests.memory sidecars Memory requests 64Mi
pgBackRest.sidecars.limits.cpu sidecars CPU limits 50m
pgBackRest.sidecars.limits.memory sidecars Memory limits 128Mi
pgBackRest.s3.enabled Enables the s3 repo backups false
pgBackRest.s3.createS3Secret Creates the s3 secret based on key and keySecret true
pgBackRest.s3.s3Secret The secret name to be created or read from s3-pgbackrest
pgBackRest.s3.s3Path The path inside the bucket where the backups will be saved to, set it to / to use the root of the bucket. /dbbackup
pgBackRest.s3.s3UriStyle Style of URL to use for S3 communication. More Info path
pgBackRest.s3.bucket The bucket to use for backups bucketName
pgBackRest.s3.endpoint The endpoint to use, for example s3.ca-central-1.amazonaws.com endpointName
pgBackRest.s3.region The region to use, not necessary if your S3 system does not specify one ca-central-1
pgBackRest.s3.key The key to use to access the bucket. MUST BE KEPT SECRET s3KeyValue
pgBackRest.s3.keySecret The key secret for the key set above. MUST BE KEPT SECRET s3SecretValue

Patroni

Patroni docs Crunchy Patroni docs

Parameter Description Default
patroni.postgresql.pg_hba pg_hba permissions "host all all 0.0.0.0/0 md5"
crunchyImage Crunchy Postgres image ...crunchy-postgres:ubi8-14.7-0
patroni.parameters.shared_buffers The number of shared memory buffers used by the server 16MB
patroni.parameters.wal_buffers The number of disk-page buffers in shared memory for WAL 64KB
patroni.parameters.min_wal_size The minimum size to shrink the WAL to 32MB
patroni.parameters.max_wal_size Sets the WAL size that triggers a checkpoint 64MB
patroni.parameters.max_slot_wal_keep_size Sets the maximum WAL size that can be reserved by replication slots 128MB

pgBouncer

A lightweight connection pooler for PostgreSQL

pgBouncer site Crunchy Postgres pgBouncer docs

Parameter Description Default
proxy.pgBouncer.image Crunchy pgBouncer image
proxy.pgBouncer.replicas Number of replicas 2
proxy.pgBouncer.requests.cpu CPU requests 1m
proxy.pgBouncer.requests.memory Memory requests 64Mi
proxy.pgBouncer.limits.cpu CPU limits 50m
proxy.pgBouncer.limits.memory Memory limits 128Mi

PG Monitor

Crunchy Postgres PG Monitor docs

Parameter Description Default
pgmonitor.enabled Enable PG Monitor (currently only PG exporter) false
pgmonitor.exporter.requests.cpu PG Monitor CPU requests 1m
pgmonitor.exporter.requests.memory PG Monitor Memory requests 64Mi
pgmonitor.exporter.limits.cpu PG Monitor CPU limits 50m
pgmonitor.exporter.limits.memory PG Monitor Memory limits 128Mi

Postgres Exporter

A Prometheus exporter for PostgreSQL

Postgres Exporter

Parameter Description Default
pgmonitor.exporter.image Crunchy PG Exporter image
pgmonitor.exporter.requests.cpu CPU requests 1m
pgmonitor.exporter.requests.memory Memory requests 64Mi
pgmonitor.exporter.limits.cpu CPU limits 50m
pgmonitor.exporterr.limits.memory Memory limits 128Mi

Data Restore CronJob

This feature allows you to set up a daily CronJob that restores data from a source S3 repository (e.g., from another database instance) into the current PostgreSQL cluster. This is useful for change data capture scenarios where you need to regularly sync data from a source database. The configuration reuses the same structure as dataSource and pgBackRest.s3 for consistency.

Configuration

Parameter Description Default
dataRestore.enabled Enable the data restore CronJob false
dataRestore.schedule Cron schedule for the restore job "0 2 * * *"
dataRestore.image pgBackRest image to use for restore crunchy-pgbackrest
dataRestore.secretName K8s secret containing S3 credentials (reuse existing) s3-pgbackrest
dataRestore.repo.name Repository name (repo1, repo2, etc.) repo2
dataRestore.repo.path S3 path prefix /habackup
dataRestore.repo.s3.bucket Source S3 bucket name bucketName
dataRestore.repo.s3.endpoint S3 endpoint URL Object store endpoint
dataRestore.repo.s3.region S3 region not-used
dataRestore.repo.s3.uriStyle S3 URI style (path or host) path
dataRestore.stanza pgBackRest stanza name db
dataRestore.target.clusterName Target cluster name (defaults to current cluster) ""
dataRestore.target.database Target database name postgres
dataRestore.resources.requests.cpu CPU requests for restore job 100m
dataRestore.resources.requests.memory Memory requests for restore job 256Mi
dataRestore.resources.limits.cpu CPU limits for restore job 500m
dataRestore.resources.limits.memory Memory limits for restore job 512Mi
dataRestore.successfulJobsHistoryLimit Number of successful jobs to keep in history 3
dataRestore.failedJobsHistoryLimit Number of failed jobs to keep in history 1
dataRestore.restartPolicy Pod restart policy for failed jobs OnFailure
dataRestore.additionalArgs Additional pgbackrest arguments []

Usage Example

The configuration reuses existing S3 secrets and follows the same patterns as dataSource:

dataRestore:
  enabled: true
  schedule: "0 2 * * *"  # Daily at 2 AM
  # Reuse existing S3 secret from dataSource or pgBackRest.s3
  secretName: "dev-s3-pgbackrest"
  repo:
    name: repo2
    path: "/habackup-source-database"
    s3:
      bucket: "source-database-backups"
      endpoint: "https://sector.objectstore.gov.bc.ca"
      region: "not-used"
      uriStyle: "path"
  stanza: db
  target:
    database: "myapp"
  additionalArgs:
    - "--log-level-console=debug"
    - "--process-max=2"

Important Notes

  • The restore uses --delta mode, which only restores changed files for efficiency
  • Reuses existing S3 secrets from dataSource or pgBackRest.s3 configuration
  • The job runs with the specified S3 repository as the source
  • Ensure the source S3 repository contains valid pgBackRest backups
  • The target cluster must be accessible and have proper credentials
  • Monitor CronJob logs for restore status and any errors
  • Configuration follows the same patterns as dataSource for consistency