Procedure for deleting a storage node VM (Compute Engine instance) that is not included in VSP One SDS Block configuration information (Cloud for Google Cloud)

Virtual Storage Platform One SDS Block and SDS Cloud Troubleshooting Reference

Version

1.19.x

Audience

anonymous

Part Number

MK-24VSP1SDS029-03

ft:lastEdition

2026-04-07

If you are instructed to do so in the flow for troubleshooting failures, perform the procedure shown further on in this section to delete a storage node VM (Compute Engine instance) that is not included in VSP One SDS Block configuration information.

CAUTION:

When you delete storage node VMs, do not delete them from the Google Cloud console or gcloud CLI. If they are deleted from the Google Cloud console or gcloud CLI, Terraform templates and cloud resources might become inconsistent, and maintenance operations might not work properly.
If you delete storage node VMs from the Google Cloud console or gcloud CLI, take action according to Action to be taken when differences occur between Terraform templates and created resources (Cloud for Google Cloud).

See Exporting configuration files (Cloud for Google Cloud) in the VSP One SDS Block and SDS Cloud System Administration to obtain configuration files from VSP One SDS Block for deleting storage nodes not included in the configuration.

Do not specify exportFileType when exporting configuration files. This step is mandatory and ensures that you use the latest configuration files.
See Deploying VM configuration files in the Terraform working directory (Cloud for Google Cloud) in the VSP One SDS Block and SDS Cloud System Administration to unzip and place the configuration file.

CAUTION:

Make sure you do not change the unzipped configuration files unless instructed to do so by the document. If you change these files, a problem such as storage cluster blockage might occur.

Verify the impact of deploying configuration files for deleting storage nodes not included in the configuration.

Run the following command.

terraform plan

When you run the command, the message Terraform will perform the following actions appears, followed by the contents of resource changes. Verify that the following resource changes are performed.

No.	Resource type	Resource operation	Resource name¹	Ponts to confirm
1	google_compute_disk.user_disks[ ]	delete	<clusterName>-snXX-userdatadiskYY	Resources for "the number of storage nodes to be deleted simultaneously² × the number of drives per node" are deleted.
2	google_compute_attached_disk.attached_user_disk[ ]	delete	-	Resources for "the number of nodes to be deleted simultaneously² × the number of drives per node" are deleted.
3	google_compute_instance.instance[ ]	delete	<clusterName>-snXX	Resources for "the number of nodes to be deleted simultaneously²" are deleted. XX of the resource name must start with the succeeding number of the number of the storage node that exists in the storage cluster.
4	google_compute_instance.instance[ ]	update in-place	<clusterName>-snXX	No confirmation is necessary. If a change is made in the following items, there is no problem. metadata deletion_protetion
5	google_compute_resource_policy.spp[ ]	delete	<clusterName>-resource-policyZZ	For Single-AZ (Single-Zone) configuration One resource must be deleted. Combination of the number of the deleted storage node XX (XX in no.4) and ZZ of the resource name must be either of the following: Data protection method: Duplication Combination of {XX, ZZ}: {01-03, 01}, {04-06, 02}, {07-09, 03}, {10-12, 04), {13-15, 05}, {16-18, 06} Data protection method: 4D+2P Combination of {XX, ZZ}: {01-03, 01}, {04-06, 02}, {07-09, 03}, {10-12, 04), {13-15, 05}, {16-18, 06} For Multi-AZ (Multi-Zone) configuration Resources must be deleted for the following numbers: 2 resources (When XX of the storage node (XX in no.4) is 16 or 17) 0 resource (When XX of the storage node (XX in no.4) is other than 16 and 17) When two resources are deleted, XX of the resource name must be 03 and 04.
1. XX, YY, and ZZ in resource names indicate the following. XX is the number assigned to the storage node. A two-digit integer starting with 01 is assigned to each of the storage nodes in the order they were configured. YY is the number assigned to the user drive of each storage node. ZZ is the number assigned to each spread placement policy. 2. The number of storage nodes to be deleted simultaneously is as follows: Zone configuration: Single-AZ(Single-Zone) Data protection method: Duplication The number of storage nodes to be deleted simultaneously: 3 Zone configuration: Single-AZ(Single-Zone) Data protection method: 4D+2P The number of storage nodes to be deleted simultaneously: 6 Zone Configuration: Multi-AZ(Multi-Zone) Data protection method: Duplication The number of storage nodes to be deleted simultaneously: 2

If there is no problem as a result of confirmation by using the terraform plan command, release deletion protection of all the storage node VMs (Compute Engine instances) confirmed for deletion in step 3.
Run the following command the same number of times as the number of the Compute Engine instances to be deleted by changing the Compute Engine instance names.
```
gcloud compute instances update <clusterName>-sn<XX> \
--zone <Zone> \
--no-deletion-protection
```
<clusterName>-sn<XX> is the name of the Compute Engine instance. <clusterName> indicates the storage cluster name and <XX> indicates the number of any Compute Engine instance confirmed for deletion in step 3.
To change the storage node resource, run the following command.
```
terraform apply
```
When Enter a value: is displayed after running the command, enter [yes].

After that, Outputs is output. After the terraform apply command is completed, the cloud resource change is completed.
If you are also using IPv6 as a communication protocol for the compute port (dual stack), unreserve the static internal IPv6 address used by the deleted storage node.
```
gcloud compute addresses delete <clusterName>-sn<XX>-compute-ipv6 \
--region <region>
```
<clusterName> indicates the storage cluster name and <XX> indicates the number of any VM resource you deleted in step 5.

When Do you want to continue (Y/n)? is displayed after running the command, enter [Y].

Then, the static IPv6 address is released from reservation.

Perform this procedure the same number of times as the number of deleted storage nodes.

Note:
With the gcloud compute addresses delete command, you can release reservation of multiple static internal IPv6 addresses simultaneously by specifying them separated by spaces.
Verify that the static IPv6 addresses have been released from reservation.
```
gcloud compute addresses list \
--filter="name~^<clusterName>-sn[0-9]{2}-compute-ipv6$"
```
Verify that the execution results of the preceding command do not include resources with the same <XX> as that of the deleted storage nodes.

The resource name is <clusterName>-sn<XX>-compute-ipv6.