What Should I Do If a Node Pool Scale-Out Fails?

Troubleshooting

Locate the fault based on the events of the failure to scale out a node pool, as shown in Table 1.

Unavailable Node Flavor

The following error message is displayed:

No available flavors for nodes.

The selected node flavor is unavailable. As a result, the node cannot be created.

Solution

1. Log in to the CCE console and click the cluster name to access the cluster console.
2. Choose Nodes in the navigation pane and click the Node Pools tab on the right.
3. Locate the row containing the node pool that failed to scale out and click Update. On the displayed Update Node Pool page, configure the parameters.
4. Update the node flavor selected for the node pool: select another available flavor.
5. After the configuration is complete, click OK.

Unavailable EVS Disk Specifications

The following error message is displayed:

No available node volumes for nodes.

The selected EVS disk specifications are unavailable. As a result, the node cannot be created.

Solution

1. Log in to the CCE console and click the cluster name to access the cluster console.
2. Choose Nodes in the navigation pane and click the Node Pools tab on the right.
3. Locate the row containing the node pool that failed to scale out and click Update. On the displayed Update Node Pool page, configure the parameters.
4. Update the storage configuration of the node pool: change the type of the system disk or data disk.
5. After the configuration is complete, click OK.

Failed to Obtain the Key Pair Used for Logging In to a Node Pool

If a node pool scale-out fails, the event contains Ecs.0314. This error code indicates that the key pair used for logging in to the node pool cannot be obtained, which results in the creation failure of a new ECS.

...call fsp to query keypair fail, error code : Ecs.0314, reason is : the keypair *** does not match the user_id ***...

The possible causes are as follows:

• The key pair selected for logging in to the node pool has been deleted.
• The key pair selected for logging in to the node pool is a private one which cannot be used by the current user to log in to the node pool and create nodes in the node pool.

Solution

• If the scale-out fails due to the first cause, you can create a key pair and then create a node pool that can be logged in to using this key pair.
• If the scale-out fails due to the second cause, only the user who created the private key pair can scale out the node pool. You can use another key pair when creating a new node pool.

Invalid KMS Key ID

When a node pool fails to be expanded, the reported event contains Ecs.0912.

{"error":{"message":"encrypted key id [***] is invalid.","code":"Ecs.0912"}}

The possible causes are as follows:

• The KMS key ID entered during node pool creation does not exist.
• The KMS key ID entered during node pool creation is the key of another user, but the user has not authorized you.

Solution

• If the scale-out fails due to the first cause, ensure that the entered key ID exists.
• If the scale-out fails due to the second cause, use the ID of the shared key that has been authorized to you.

The Security Group Specified by the Node Pool Deleted

When a node pool fails to scale out, the event contains the following information:

Security group [*****] not found

Or

{"error":{"message":"{\"NeutronError\":{\"detail\":\"\",\"message\":\"Security group ***** does not exist\",\"type\":\"SecurityGroupNotFound\"}}","code":"Ecs.0005"}}

This issue can arise in the following scenarios:

• Scenario 1: A custom security group is set up for the node pool but gets deleted, so the node pool scale-out fails.
• Scenario 2: No custom security group is configured for the node pool and the default security group is deleted, so the node pool scale-out fails.

Solution

• Scenario 1: Update the security group specified by the customSecurityGroups field by calling the API for updating a node pool. For details, see Updating a Specified Node Pool.
• Scenario 2: Log in to the CCE console and change the default node security group on the Settings page of the cluster. The new node security group must meet the communication rules of the cluster ports. For details, see How Do I Modify Cluster Security Group Rules?

The Enterprise Project Specified by the Node Pool Deleted

When a node pool fails to scale out, the event contains the following information:

The Enterprise Project ID [*****] not exist

The enterprise project configured for the node pool has been deleted from EPS.

Solution

1. Log in to the CCE console and click the cluster name to access the cluster console.
2. Choose Nodes in the navigation pane and click the Node Pools tab on the right.
3. Locate the row containing the node pool that failed to scale out and click Update. On the displayed Update Node Pool page, configure the parameters.
4. Select another enterprise project for the node pool.
5. After the configuration is complete, click OK.

The Subnet Specified by the Node Pool Deleted

When a node pool fails to scale out, the event contains the following information:

Subnet [*****] not found;CloudService=CCE,Service=S-CCE

The subnet associated with the node pool has been deleted.

• If the node pool is associated with multiple subnets and one is deleted, scale-out skips that subnet and continues provisioning nodes in the remaining subnets. However, if the remaining subnets have insufficient available IP addresses, scale-out may still fail.
• If the node pool is associated with only one subnet, scale-out fails continuously because the subnet is unavailable after deletion.

Solution

1. Log in to the CCE console and click the cluster name to access the cluster console.
2. Choose Nodes in the navigation pane and click the Node Pools tab on the right.
3. Locate the row containing the node pool that failed to scale out and click Update. On the displayed Update Node Pool page, configure the parameters.
4. Update the subnet associated with the node pool.
5. After the configuration is complete, click OK.

The Image Specified by the Node Pool Deleted

When a node pool fails to scale out, the event contains the following information:

...{error":{message":"query image [***] fail:[404 Not Found\n\nNo image found with ID ***\n\n ]","code":"Ecs.0304"}}...

The image specified for the node pool has been deleted from IMS.

Solution

1. Log in to the CCE console and click the cluster name to access the cluster console.
2. Choose Nodes in the navigation pane and click the Node Pools tab on the right.
3. Locate the row containing the node pool that failed to scale out and click Update. On the displayed Update Node Pool page, configure the parameters.
4. Update the image specified for the node pool.
5. After the configuration is complete, click OK.