Facing error "metric pg_database_size_bytes was collected before with the same name and label values"

What did you do?
We deployed postgres exporter version 0.11.1 on Azure using the official community helm chart. We are facing an issue with the below query from the chart:

pg_database: query: "SELECT pg_database.datname, pg_database_size(pg_database.datname) as size_bytes FROM pg_database" master: true cache_seconds: 30 metrics: - datname: usage: "LABEL" description: "Name of the database" - size_bytes: usage: "GAUGE" description: "Disk space used by the database"

We keep observing below error on the /metrics endpoint for the above query:

An error has occurred while serving metrics:

7 error(s) occurred:
* collected metric "pg_database_size_bytes" { label:<name:"datname" value:"azure_maintenance" > label:<name:"server" value:"<IP ADDRESS>:5432" > gauge:<value:8.401455e+06 > } was collected before with the same name and label values
* collected metric "pg_database_size_bytes" { label:<name:"datname" value:"template1" > label:<name:"server" value:"<IP ADDRESS>:5432" > gauge:<value:8.159747e+06 > } was collected before with the same name and label values
* collected metric "pg_database_size_bytes" { label:<name:"datname" value:"postgres" > label:<name:"server" value:"<IP ADDRESS>:5432" > gauge:<value:8.589871e+06 > } was collected before with the same name and label values
* collected metric "pg_database_size_bytes" { label:<name:"datname" value:"template0" > label:<name:"server" value:"<IP ADDRESS>:5432" > gauge:<value:8.159747e+06 > } was collected before with the same name and label values
* collected metric "pg_database_size_bytes" { label:<name:"datname" value:"azure_sys" > label:<name:"server" value:"<IP ADDRESS>:5432" > gauge:<value:8.974895e+06 > } was collected before with the same name and label values
* collected metric "pg_database_size_bytes" { label:<name:"datname" value:"testdb1" > label:<name:"server" value:"<IP ADDRESS>:5432" > gauge:<value:8.311343e+06 > } was collected before with the same name and label values
* collected metric "pg_database_size_bytes" { label:<name:"datname" value:"example" > label:<name:"server" value:"<IP ADDRESS>:5432" > gauge:<value:3.0814767e+07 > } was collected before with the same name and label values

What did you expect to see?
We expect to see regular metrices for pg_database_size_bytes with database name and size in labels

What did you see instead? Under which circumstances?
We keep seeing the error listed above on the /metrics endpoint

Environment
Azure

• postgres_exporter version: 0.11.1

• postgres_exporter flags: None

• PostgreSQL version: 13.7

• Logs: Found nothing specific in the logs.

Start at the /metrics endpoint and trace handling of the pg_database query configured in the community Helm chart. Reproduce with postgres_exporter 0.11.1, PostgreSQL 13.7, and the shown query, then inspect why identical pg_database_size_bytes samples are collected. Done means the endpoint exposes one valid metric series per database without duplicate-collection errors.