FusionAuth/fusionauth-containers

[helm] pods stuck in init using custom PostgreSQL and Elasticsearch

Closed this issue · 1 comments

nodesocket commented

Using my own Elasticsearch and PostgreSQL. So in values.yaml I am setting:

  ...
  data:
    memory: 256M
    database:
      # if empty {{- .Release.Name -}}-postgresql will be used
      host: postgres-master.default.svc.cluster.local
      port: 5432
      tls: false
      name: fusionauth
      user: localhost
      password: localhost
      root:
        user: localhost
        password: localhost
    elasticsearch:
      host: elasticsearch-master.default.svc.cluster.local
      port: 9200
  ...
  elasticsearch:
    enabled: false

  postgresql:
    enabled: false

But the pods are just stuck on:

fusionauth-6c4745bfd9-6jgth     0/1     Init:0/2           0          16m
fusionauth-6c4745bfd9-r48k2     0/1     Init:0/2           0          16m
fusionauth-6c4745bfd9-sjl72     0/1     Init:0/2           0          16m

Looking at the pods ENVARS I see:

    Environment:
      DATABASE_USER:              localhost
      DATABASE_PASSWORD:          localhost
      DATABASE_ROOT_PASSWORD:     localhost
      DATABASE_ROOT_USER:         localhost
      DATABASE_URL:               jdbc:postgresql://fusionauth-postgresql:5432/fusionauth
      FUSIONAUTH_SEARCH_SERVERS:  http://elasticsearch-master:9200
      FUSIONAUTH_MEMORY:          256M

Is that right? I would have expected to see the custom hosts in ENVARs

drpebcak commented

@nodesocket can you try this with the latest version of the chart from here: https://github.com/FusionAuth/charts

Some things have changed, so you will have to modify your values, but it should work.

  • 👍1