Vault metrics serverd only by one node out of 3

Hi, so yesterday i finally enabled telemetry to gather metrics from vault.
I have 2 clusters in my k8s setup, one for autounsealing divided by namespace, here everything works. I can get most of the metrics i’m interested in.
However in the second cluster which has identical configuration in terms of metrics/telemetry config. I get response only from one of the pods. The other ones end the request with message:

curl -k https://vault.services-dev.svc:8200/v1/sys/metrics\?format=prometheus
prometheus is not enabled

Can’t really figure out what is happening.
Configs from both vault clusters are here
Not working cluster:

      config: |
        ui = true
        listener "tcp" {
          tls_disable = 0
          address = "0.0.0.0:8200"
          cluster_address = "0.0.0.0:8300"
          tls_cert_file = "/vault/userconfig/tls-server/tls.crt"
          tls_key_file = "/vault/userconfig/tls-server/tls.key"
          tls_min_version = "tls12"
          telemetry {
            unauthenticated_metrics_access = "true"
          }
        }
        seal "transit" {
          address = "https://vault-auth.vault-auth.svc:8200"
          # please add correct token to the config otherwise it will not work!!
          # this needs to be the unwrapped one !!
          token = "unique_token"
          disable_renewal = "false"
          key_name = "autounseal"
          mount_path = "transit/"
          tls_skip_verify = "true"
        }
        storage "postgresql" {
          connection_url = "postgres://user:pass@vault-db-postgresql-ha-pgpool.services-dev.svc:5432/vault?sslmode=disable"
          table = "vault_kv_store"
          ha_enabled = "true"
          ha_table = "vault_ha_locks"
        }
        telemetry {
          disable_hostname = true
          prometheus_retention_time = "72h"
        }
        service_registration "kubernetes" {}
        disable_mlock = true

Working cluster:

      config: |
        ui = true
        listener "tcp" {
          tls_disable = 0
          address = "0.0.0.0:8200"
          cluster_address = "0.0.0.0:8300"
          tls_cert_file = "/vault/userconfig/tls-server/tls.crt"
          tls_key_file = "/vault/userconfig/tls-server/tls.key"
          tls_min_version = "tls12"
          telemetry {
            unauthenticated_metrics_access = "true"
          }
        }
        storage "postgresql" {
          connection_url = "postgres://user:pass@vault-db-postgresql-ha-pgpool.services-dev.svc:5432/vault_auth?sslmode=disable"
          table = "vault_kv_store"
          ha_enabled = "true"
          ha_table = "vault_ha_locks"
        }
        telemetry {
          disable_hostname = true
          prometheus_retention_time = "72h"
        }

Maybe found out the problem. I removed the cluster_address config line. Downscaled to 1 replica and rescaled to 5 and back to 3 when i was testing this. Suddenly it works. But still it is black magick to me