Problems with Restart

You might experience unexpected side-effects after restarting an Oracle Java Cloud Service instance or individual VMs. These effects can also occur after patching, which restarts VMs.

Restart fails after a scale down operation intended to remedy a quota breach

You can scale down a Oracle Database Cloud Service database deployment or Oracle Java Cloud Service instance if you have a quota breach in your account. Scaling down reduces compute resources. However, the automatic restart action can fail after scale-down.

For example, you could scale down a node from shape oc5 to oc3. Oracle Java Cloud Service puts the service instance into Maintenance mode, changes the state of the node to Configuring, and stops any servers running on the node. After applying the changes, Oracle Java Cloud Service is supposed to start the servers automatically. If the quota breach is not cleared by the time the orchestration is restarted with the smaller shape, the automatic server restart action could fail.

If the restart action fails, wait one hour for the quota breach to clear, then restart the service instance by using the Oracle Java Cloud Service Console.

Monitor a VM’s boot log

You can monitor the boot progress of individual VMs by using Oracle Cloud Infrastructure Compute Classic. See Viewing the Boot Log of an Instance in Using Oracle Cloud Infrastructure Compute Classic. Ignore information in this topic about the Compute API.

My custom storage volumes have become detached

Custom storage volumes you have added after creating an Oracle Java Cloud Service instance will become detached after restart operations.

Do not attach custom storage volumes to a service instance’s VMs. Any custom storage volumes are detached if the service instance is restarted.

If a service instance requires additional storage, add storage by scaling the service instance’s nodes as explained in Scale an Oracle Java Cloud Service Node.

My content changes on the Boot/OS volume are gone

The Boot/OS volume of any service instance provisioned before the mid-August 2015 update to Oracle Java Cloud Service is an ephemeral disk volume. Content added to an ephemeral Boot/OS volume does not persist if the service instance is restarted. To avoid the risk of data loss, do not add content to an ephemeral Boot/OS volume.

This restriction does not apply to more recent service instances. The Boot/OS volume of any service instance provisioned after the mid-August 2015 update to Oracle Java Cloud Service is persistent. Content added to a persistent Boot/OS volume is retained if the service instance is restarted.

For details of this volume, see About the Storage Volumes Attached to the WebLogic Server Nodes.

Restart fails with the Out of host capacity error message

When restart fails due to 'Out of host capacity', the JCS Activity Log displays the following error message:

Unable to start the following Compute resources: Compute Error: Error returned by InstanceAction operation in Compute service.(500, InternalError, false) Out of host capacity.

Check if the JCS instance is using a shape from the previous generation of shapes. For information about shapes, see Compute Shapes.

If yes, Oracle recommends to resize to a newer generation shape. See Resize a Service Instance.