Skip to main content

Troubleshooting — Redis

The cluster stays in "Creating" status​

Cause: the Redis nodes, the Sentinels and their volumes are still being provisioned.

Solution:

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

"An error occurred while updating the cluster." when saving​

Cause: the change requests a version lower than the current one (for example 7 on a cluster running 8), or a Volume Size (GB) smaller than the current size. These changes are refused, and the error message includes the reason returned by the API. On the API side, these refusals carry the codes 120203 (version) and 120204 (volume).

Solution: keep the current version or a higher one, and a volume size at least equal to the current size.

The host shows "Waiting for allocation..."​

Cause: the public network is disabled, or the public IP address has not been assigned yet.

Solution:

  1. Open Edit and check the External access option. Enable it if you need to connect from the Internet, then click Save changes.
  2. Wait a few moments and refresh the cluster page.

Connection timeout​

Cause: the address or port used is incorrect, the cluster is not ready, or a firewall is blocking port 6379.

Solution:

  1. Check that the Status in the Connection section is Ready.
  2. Copy the Host with the copy button to avoid typing errors.
  3. Check that no outbound firewall on your network is blocking port 6379.

Authentication fails (NOAUTH or WRONGPASS)​

Cause: the client does not send a password, uses a wrong password, or a password revoked by a rotation.

Solution:

  1. Check the value passed to the client (REDISCLI_AUTH, -a option or application configuration).
  2. If in doubt, generate a new password from the Security section (Rotate password) and update your applications. See Renew the password.
  3. If the Authentication required option has been changed, update the clients accordingly.

Memory saturated (OOM command not allowed)​

Cause: the dataset exceeds the memory allocated by the preset.

Solution:

  1. Check memory usage: 
    redis-cli -h <host> -p 6379 INFO memory
  2. Switch to a larger Preset via Edit. See Change resources.
  3. If Redis is used as a cache, set expiration times (EXPIRE) on your keys to limit dataset growth.

Failover does not happen​

Cause: the cluster has only one replica; no replica can be promoted to master.

Solution: the number of replicas cannot be changed after creation. Create a new cluster with at least 2 replicas (3 in production) and migrate your data, or contact support. See Configure high availability.