[bitnami/postgresql] password authentication failed for a user/pwd specified using values: "auth.username" and "auth.password"

[cmoulliard]

Name and Version

bitnami/postgresql 11.9.13

What architecture are you using?

amd64

What steps will reproduce the bug?

1. Create a kind cluster
2. Deploy the helm chart as such

DB_USERNAME=healthy
DB_PASSWORD=healthy

VERSION=11.9.13
helm install postgresql bitnami/postgresql \
  --version $VERSION \
  --set auth.username=$DB_USERNAME \
  --set auth.password=$DB_PASSWORD \
  --set auth.database=fruits_database \
  --create-namespace \
  -n db

3. Check if the pod has been well created

kubectl get po -n db
NAME           READY   STATUS    RESTARTS   AGE
postgresql-0   1/1     Running   0          18s

4. Access the DB using the psql client

export POSTGRES_PASSWORD=$(kubectl get secret --namespace db postgresql -o jsonpath="{.data.password}" | base64 -d)
kubectl run postgresql-client --rm --tty -i --restart='Never' --namespace db --image docker.io/bitnami/postgresql:14.5.0-debian-11-r35 --env="PGPASSWORD=$POSTGRES_PASSWORD" \
      --command -- psql --host postgresql -U healthy -d fruits_database -p 5432
psql: error: connection to server at "postgresql" (10.96.93.222), port 5432 failed: FATAL:  password authentication failed for user "healthy"
pod "postgresql-client" deleted
pod db/postgresql-client terminated (Error)

5. Verify the password matches what it has ben passed as helm value

echo $POSTGRES_PASSWORD
healthy

As the password encoded corresponds to what it has been passed as value to the helm chart using auth.password, there is certainly an issue with the init step as we cannot access the db fruits_database using the user healthy and password healthy

Are you using any custom parameters or values?

--set auth.username=$DB_USERNAME \
--set auth.password=$DB_PASSWORD \
--set auth.database=fruits_database \

What is the expected behavior?

I should be able to auth the user using the password defined as parameter

What do you see instead?

psql: error: connection to server at "postgresql" (10.96.93.222), port 5432 failed: FATAL: password authentication failed for user "healthy"

Additional information

No response

[javsalgar]

Hi!

Thanks for opening the ticket. Could you change the title to something more descriptive?

[cmoulliard]

The problem I raised here hs been discussed part of this blog post here: https://medium.com/@hijessicahsu/deploy-postgres-on-minikube-5cd8f9ffc9c.

It is related to the fact that if a postgres instance has been created before, data still persist within the PVC and are not cleaned. If I delete the PVC, then I can access the DB with the user/password defined

kubectl run postgresql-client --rm --tty -i --restart='Never' --namespace db --image docker.io/bitnami/postgresql:14.5.0-debian-11-r35 --env="PGPASSWORD=$POSTGRES_PASSWORD" --command -- psql --host postgresql -U healthy -d fruits_database -p 5432
If you don't see a command prompt, try pressing enter.

fruits_database=>

[javsalgar]

Hi,

Yes, the issue is that PVC from StatefulSets don't get removed by default. Therfore letfovers from previous installations should be removed.

[cmoulliard]

This is why I propose 2 changes within the documentation here: https://github.com/bitnami/charts/tree/main/bitnami/postgresql#uninstalling-the-chart

1. Rephrase

"The command removes all the Kubernetes components but PVC's associated with the chart and deletes the release." -->
"The command removes all the Kubernetes components except the PVC's associated with the chart and deletes the release."

2. Clearly warn the user

I propose to also add a note within the section to uninstall the helm chart and tell to the user that it is mandatory to delete the PVC when they will re-install the chart using a different database name, user and password !

[javsalgar]

Hi,

Thanks for the input! Would you like to submit a PR updating the README?

[github-actions]

This Issue has been automatically marked as "stale" because it has not had recent activity (for 15 days). It will be closed if no further activity occurs. Thanks for the feedback.

[github-actions]

Due to the lack of activity in the last 5 days since it was marked as "stale", we proceed to close this Issue. Do not hesitate to reopen it later if necessary.

[osirisguitar]

The problem I raised here hs been discussed part of this blog post here: https://medium.com/@hijessicahsu/deploy-postgres-on-minikube-5cd8f9ffc9c.

It is related to the fact that if a postgres instance has been created before, data still persist within the PVC and are not cleaned. If I delete the PVC, then I can access the DB with the user/password defined

kubectl run postgresql-client --rm --tty -i --restart='Never' --namespace db --image docker.io/bitnami/postgresql:14.5.0-debian-11-r35 --env="PGPASSWORD=$POSTGRES_PASSWORD" --command -- psql --host postgresql -U healthy -d fruits_database -p 5432
If you don't see a command prompt, try pressing enter.

fruits_database=>

Lifesaver, I was tearing my hair out over this :-D

[jyorko]

So are you supposed to just delete the pvc and try to login again, im not getting it, i have had this error over and over and over again, not just on Kuberenetes with this specific helm chart but with docker as well. I have pruned my docker system, deleted entire namespaces and done fresh installs with fresh passwords, and still getting the error?