Logical backup job fails if pooler is enabled for the cluster - pg_dump: error: connection to database "template1"

Please, answer some short questions which should help us to understand your problem / question better?

• Which image of the operator are you using?
  registry.opensource.zalan.do/acid/postgres-operator:v1.6.1
• **Where do you run it - cloud or metal?
  Kubernetes with kops on AWS
• Are you running Postgres Operator in production?
  we are going to
• Type of issue?
  Bug report

When the cluster manifest enables poolers cronjobs for logical backups fail with error pg_dump: error: connection to database "template1" failed: ERROR: unexpected response from login query

job full logs:

  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100 12220    0 12220    0     0   917k      0 --:--:-- --:--:-- --:--:--  994k
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100 20702    0 20702    0     0  1555k      0 --:--:-- --:--:-- --:--:-- 1555k
+ dump
+ /usr/lib/postgresql/13/bin/pg_dumpall
+ compress
+ pigz
+ upload
+ case $LOGICAL_BACKUP_PROVIDER in
++ estimate_size
++ /usr/lib/postgresql/13/bin/psql -tqAc 'select sum(pg_database_size(datname)::numeric) from pg_database;'
+ aws_upload 6710488
+ declare -r EXPECTED_SIZE=6710488
++ date +%s
+ PATH_TO_BACKUP=s3://postgres-operator-backups-bucket/spilo/test-db/aae491ea-8a1b-4d1f-b5bd-755466eae34c/logical_backups/1616113817.sql.gz
+ args=()
+ [[ ! -z 6710488 ]]
+ args+=("--expected-size=$EXPECTED_SIZE")
+ [[ ! -z '' ]]
+ [[ ! -z us-east-1 ]]
+ args+=("--region=$LOGICAL_BACKUP_S3_REGION")
+ [[ ! -z AES256 ]]
+ args+=("--sse=$LOGICAL_BACKUP_S3_SSE")
+ aws s3 cp - s3://postgres-operator-backups-bucket/spilo/test-db/aae491ea-8a1b-4d1f-b5bd-755466eae34c/logical_backups/1616113817.sql.gz --expected-size=6710488 --region=us-east-1 --sse=AES256
pg_dump: error: connection to database "template1" failed: ERROR:  unexpected response from login query
pg_dumpall: error: pg_dump failed on database "template1", exiting

cluster manifest:

apiVersion: "acid.zalan.do/v1"
kind: "postgresql"
metadata:
  name: "test-db"
  namespace: "postgres-test"
  labels:
    team: test
spec:
  teamId: "test"
  postgresql:
    version: "13"
    parameters:
      max_connections: "100"
  numberOfInstances: 2
  enableConnectionPooler: true
  enableReplicaConnectionPooler: true
  enableLogicalBackup: true
  volume:
    size: "5Gi"
  users:
    admin:
    - superuser
    - createdb
    test-user: []
  databases:
    test: test-user
  resources:
    requests:
      cpu: 100m
      memory: 500Mi
    limits:
      cpu: 1000m
      memory: 2000Mi
  sidecars:
  - name: postgres-exporter
    image: wrouesnel/postgres_exporter
    env:
    - name: DATA_SOURCE_USER
      valueFrom:
        secretKeyRef:
          name: postgres.test-db.credentials.postgresql.acid.zalan.do
          key: username
    - name: DATA_SOURCE_PASS
      valueFrom:
        secretKeyRef:
          key: password
          name: postgres.test-db.credentials.postgresql.acid.zalan.do
    - name: "DATA_SOURCE_NAME"
      value: "postgresql://$(DATA_SOURCE_USER):$(DATA_SOURCE_PASS)@$(POD_NAME):5432/postgres"
    ports:
    - containerPort: 9187
      protocol: TCP
    resources:
      limits:
        cpu: 500m
        memory: 100Mi
      requests:
        cpu: 100m
        memory: 100Mi

Start with the logical-backup job's pg_dumpall invocation and the manifest settings enableConnectionPooler, enableReplicaConnectionPooler, and enableLogicalBackup; reproduce the template1 failure with the supplied Kubernetes configuration. Done means logical backups complete successfully when poolers are enabled.