fix: Replace migrate hooks with longer timeouts and retries

[tamalsaha]

This makes it possible to use this chart with fluxcd

Description

Helm charts and Jobs don't work well together. This is evident from many prs trying to address some aspects of this issue. I try to document the issues as best I can think of here: https://x.com/tsaha/status/1805382111844778275

This pr tries to address this issue as best as possible I think.

If the hooks are used, we can't use fluxcd to install this chart with --wait flag. See here: fluxcd/flux2#1085 (comment)

What seems to work well in our experience is just run the migrate job at the same time as the potential db sub-chart, increase the timeout of the migrate job and increase number of retries. #130 pr ensures that at least one successful run of the migrate job will be considered successful. So, extending the timeout and increasing retries works well even if the db subchart takes unusally long time to be ready.

References

Review Checklist

• I have clicked on "allow edits by maintainers".
• I have added documentation for new/changed functionality in this PR or in a PR to openfga.dev [Provide a link to any relevant PRs in the references section above]
• The correct base branch is being used, if not main
• I have added tests to validate that the change in functionality is working as expected

[tamalsaha]

Tested in ArgoCD using

apiVersion: argoproj.io/v1alpha1
kind: Application
metadata:
  name: openfga
  namespace: argocd
spec:
  project: default
  source:
    chart: openfga
    repoURL: ghcr.io/appscode-charts
    targetRevision: 0.2.13
    helm:
      releaseName: openfga
      valuesObject:
        datastore:
          engine: postgres
          uri: "postgres://postgres:[email protected]:5432/postgres?sslmode=disable"
        postgresql:
          enabled: true
          auth:
            postgresPassword: password
            database: postgres
  destination:
    server: "https://kubernetes.default.svc"
    namespace: default