Skip to main content

Troubleshooting — PostgreSQL

The cluster stays in "Creating" status​

Cause: the instances and their volumes are being provisioned. This can take several minutes, longer with several replicas.

Solution:

  1. Wait a few minutes and refresh the cluster page.
  2. If the status does not change after about fifteen minutes, contact support, providing the project and the cluster name.

The cluster changes to "Error" status​

Cause: the platform could not apply the cluster's creation or its last change. After a change, the cluster may keep running with its previous configuration.

Solution: contact support, providing the project, the cluster name and the change requested.

Unable to get past the wizard's Configuration step​

Cause: the requested configuration exceeds the project quotas (CPU, memory or storage). The message "Storage quota exceeded for this project" may appear under the Disk size (GB) field.

Solution:

  1. Check the quota banner at the top of the wizard.
  2. Reduce the Instance preset, the Disk size (GB) or the Number of replicas: consumption is multiplied by the number of replicas.
  3. If the project needs additional quotas, contact support.

Connection refused or timed out​

Cause: external access is disabled, the IP address has not been assigned yet, or the client is using the wrong address or port.

Solution:

  1. On the cluster page, check that the External Access card shows Enabled. Otherwise, enable it via Edit.
  2. Check that the External zone of the Connection and Databases card contains an address, and not Disabled or "Waiting for allocation…".
  3. Use port 5432 and test connectivity: 
    pg_isready -h <host> -p 5432
  4. Check that no outbound firewall on your network blocks port 5432.

Authentication rejected (password authentication failed)​

Cause: wrong password or password revoked by a rotation, or user without rights on the target database.

Solution:

  1. In the Users tab, check that the user exists and has access to the database in use (Databases column).
  2. If necessary, add the access via Actions → Manage Access.
  3. If the password has been lost or has changed, generate a new one via Actions → Change Password, then update your applications.

Permission denied on a table (permission denied)​

Cause: the user has the Read-only right on the database, or has no access to this database.

Solution: via Actions → Manage Access, grant the Administrator (Admin) right on the database concerned, then reconnect.

Slow performance​

Cause: the allocated resources are insufficient for the load, or some queries are not optimized.

Solution:

  1. Enable the pg_stat_statements extension on the database (Actions → Manage extensions) and identify the most expensive queries: 
    SELECT query, calls, mean_exec_time
    FROM pg_stat_statements
    ORDER BY mean_exec_time DESC
    LIMIT 10;
  2. Add the missing indexes.
  3. If resources are saturated, switch to a higher preset via Edit. See Scale resources.
  4. To tune PostgreSQL parameters (shared_buffers, work_mem, max_connections), contact support: these parameters are not offered in the console.

Disk full​

Cause: the data volume has reached the Allocated Size.

Solution: increase the Disk size (GB) via Edit, within the project's storage quota. The size cannot be reduced afterwards. If needed, delete obsolete data and run VACUUM to reclaim space.