Skip to content

Upgrading

Warning

These upgrade instructions cover the upgrades of the helm chart only, they do NOT cover any steps necessary to upgrade the Nautobot application. Please see the official core documentation for further details.

Major Version Upgrades

To 3.x

  • The postgresql and redis subcharts are no longer deployed by default. They must be explicitly enabled if used.
postgresql:
  enabled: true
redis:
  enabled: true
  • The rabbitmq subchart was removed. It must be deployed as a separate installation if used.
  • The mariadb subchart was removed. If must be deployed as a separate installation if used.

  • The postgresqlha subchart was removed. If must be deployed as a separate installation if used.

  • The username field is now required in the existing secret if nautobot.db.existingSecret is used. By default, the username and password keys are used. The following snippets show the sample secret and values.yaml settings:

---
apiVersion: "v1"
kind: "Secret"
metadata:
  name: "nautobot-db-credentials"
  namespace: "nautobot"
type: "Opaque"
data:
  username: "bmF1dG9ib3Q="
  password: "cGFzc3dvcmQ="

nautobot:
  db:
    existingSecret: "nautobot-db-credentials"
  • The default image was changed to 3.x. Make sure that you explicitly set the value to 2.x if you are not ready to upgrade. Also the registry was changed from ghcr.io to registry-1.docker.io.
nautobot:
  image:
    registry: "ghcr.io"
    repository: "nautobot/nautobot"
    tag: "2.4.23-py3.11"
  • The Nautobot Celery readiness and liveness probes were changed to use files instead of celery pings. This change requires at least Nautobot version 3.0.1. If older version is used disable this feature:
celery:
  celery_health_probes_as_files: false

  livenessProbe:
    enabled: true
    exec:
      command:
        - "bash"
        - "-c"
        - "nautobot-server celery inspect ping --destination celery@$HOSTNAME"
    initialDelaySeconds: 10
    periodSeconds: 60
    timeoutSeconds: 10
    failureThreshold: 3
    successThreshold: 1

  readinessProbe:
    enabled: true
    exec:
      command:
        - "bash"
        - "-c"
        - "nautobot-server celery inspect ping --destination celery@$HOSTNAME"
    initialDelaySeconds: 10
    periodSeconds: 60
    timeoutSeconds: 10
    failureThreshold: 3
    successThreshold: 1

  • The configuration for Celery workers includes new options. If the options are currently set with environmental variables, remove those options and enable them through the Celery settings.

    • Set the value for celery.task_acks_late and remove the CELERY_TASK_ACKS_LATE environmental variable.
    • Set the value for celery.worker_prefetch_multiplier and remove the NAUTOBOT_CELERY_WORKER_PREFETCH_MULTIPLIER environmental variable.

To 2.x

Note

These instructions assume you have followed the very basic installation instructions as described above. If you have utilized existing secrets or advanced options, please see the release notes referenced below for possible additional steps.

  • The celeryWorker and celeryBeat values change to workers.default and workers.beat respectively.

  • The PostgreSQL subchart was upgraded from version 10 to version 12, this was a breaking change, see the release notes for more information. At a minimum the following values must change:

    • postgresql.postgresqlPostgresPassword changes to postgresql.auth.postgresPassword
    • postgresql.postgresqlUsername changes to postgresql.auth.username
    • postgresql.postgresqlPassword changes to postgresql.auth.password
    • postgresql.postgresqlDatabase changes to postgresql.auth.database
    • postgresql.existingSecret changes to postgresql.auth.existingSecret
    • postgresql.securityContext changes to postgresql.primary.podSecurityContext
    • postgresql.containerSecurityContext changes to postgresql.primary.containerSecurityContext

  • The Redis subchart was upgraded from version 16 to version 17, this was a breaking change, see the release notes for more information. No default values passed to this chart have changed.

  • The PostgreSQL-HA subchart was upgraded from version 8 to version 11, this was a breaking change, see the release notes for more information. At a minimum the following values must change:

    • postgresqlha.postgresql.securityContext changes to postgresqlha.postgresql.podSecurityContext
    • postgresqlha.metrics.securityContext changes to postgresqlha.metrics.podSecurityContext

  • The MariaDB subchart was upgraded from version 10 to version 11, this is a breaking change, see the release notes for more information. No default values passed to this chart have changed.

Example Upgrade Procedure

Warning

This is only an example based on the most simple default deployment case. Be sure to read the Bitnami PostgreSQL Release Notes before upgrading.

  1. Please verify if any values in your local values.yaml file match the above values you will need to migrate them. If you do not have the original values.yaml the chart was deployed with you can use helm get values -n $NAMESPACE -o yaml to get the deployed values.
  2. ALWAYS BACKUP YOUR DATABASE BEFORE UPGRADING!

  3. Run the following Commands:

    Note

    These steps were derived from the Bitnami Release notes from Upgrading PostgreSQL, be sure to check them for any deltas you may have.

    export NAMESPACE=nautobot  # Be sure to use the correct namespace here
export POSTGRES_SECRET_NAME=nautobot-postgresql  # If you have changed the default make sure you change it here
export STATEFULSET_NAME=nautobot-postgresql  # If you have changed the default make sure you change it here
export POSTGRESQL_PASSWORD=$(kubectl get secret --namespace $NAMESPACE $POSTGRES_SECRET_NAME -o jsonpath="{.data.postgresql-password}" | base64 --decode)
export POSTGRESQL_POSTGRES_PASSWORD=$(kubectl get secret --namespace $NAMESPACE $POSTGRES_SECRET_NAME -o jsonpath="{.data.postgresql-postgres-password}" | base64 --decode)
export POSTGRESQL_PVC=$(kubectl get pvc --namespace $NAMESPACE -l app.kubernetes.io/instance=nautobot,app.kubernetes.io/name=postgresql,role=primary -o jsonpath="{.items[0].metadata.name}")
export CURRENT_VERSION=$(kubectl exec --namespace $NAMESPACE nautobot-postgresql-0 -- bash -c 'echo $BITNAMI_IMAGE_VERSION')

echo $POSTGRESQL_PASSWORD
echo $POSTGRESQL_POSTGRES_PASSWORD
echo $POSTGRESQL_PVC
echo $CURRENT_VERSION

  4. Ensure the values output above are correct.

  5. Update your values.yaml moving the values to the new names as described above and in the Bitnami PostgreSQL Release Notes](https://docs.bitnami.com/kubernetes/infrastructure/postgresql/administration/upgrade/). A basic example would take a simple values file such as:

    celeryBeat:
  resources:
    requests:
      cpu: "1000m"
      memory: "2G"
celeryWorker:
  replicaCount: 4
postgresql:
  postgresqlPassword: change-me
redis:
  auth:
    password: change-me

    and it would change to:

    Note

    Make sure to replace the postgresPassword, tag, and existingClaim with the values from above.

    workers:
  beat:
    resources:
      requests:
      cpu: "1000m"
      memory: "2G"
  default:
    replicaCount: 4
postgresql:
  auth:
    password: change-me
    postgresPassword: ${POSTGRESQL_POSTGRES_PASSWORD from above}
  image:
    tag: ${CURRENT_VERSION from above}
  primary:
    persistence:
      existingClaim: ${POSTGRESQL_PVC from above}
redis:
  auth:
    password: change-me

    Warning

    The following steps are disruptive and should be executed during a maintenance window.

  6. Scale the Nautobot services to 0

    kubectl scale --namespace $NAMESPACE --replicas 0 deployment/nautobot deployment/nautobot-celery-beat deployment/nautobot-celery-worker

  7. This would be a good time to get a pristine backup of your database

    kubectl delete statefulsets.apps $STATEFULSET_NAME --namespace $NAMESPACE --cascade=orphan
kubectl delete secret $POSTGRES_SECRET_NAME --namespace $NAMESPACE
kubectl delete pod --namespace default nautobot-postgresql-0
helm repo update nautobot
helm upgrade nautobot nautobot/nautobot -f ${New values.yaml file from above}

Note

The instructions above reuse the same PostgreSQL version you were using in your chart release. Otherwise, you will find an error such as the one below when upgrading since the new chart major version also bumps the application version. To workaround this issue you need to upgrade database, please refer to the official PostgreSQL documentation for more information about this.

Minor Version Upgrades

Nautobot Versions

This chart doesn't always get a new release with a new release of Nautobot, therefore the chart versions do NOT correspond to a Nautobot version. It is possible to deploy newer Nautobot versions with this chart however, to do this simply run a command similar to:

helm upgrade nautobot nautobot/nautobot --reuse-values --set nautobot.image.tag=1.5.X-py3.10

1.0.x to 1.1.x

Following the normal helm upgrade procedures is sufficient for upgrading during this release:

helm repo update nautobot
helm upgrade nautobot nautobot/nautobot --reuse-values