$ gcloud alpha compute instance-groups managed instance-configs \
    update example-group --instance=example-instance \
    --update-stateful-disk=device-name=my-disk-3,\
    source=projects/my-project/zones/us-central1-a/disks/my-disk-3 \
    --remove-stateful-disks=my-disk-1,my-disk-2

POSITIONAL ARGUMENTS

NAME

Name of the managed instance group to update per instance config for.

REQUIRED FLAGS

--instance=INSTANCE

URI to existing or non existing instance.
Name - last part of URI - will be preserved for existing per instance configs.
For zonal managed instance groups there is no need to specify the whole URI to the instance - for this case instance name can be applied instead of URI.

OPTIONAL FLAGS

--force-instance-update

The changes will be applied immediately to the instances. If this flag is not provided, the changes will be applied once the instances are restarted or recreated.
Example: let's say we have an instance with a disk attached to it and an override for the disk. If we decide to delete the override and provide this flag, this will instantly recreate the instance and detach the disk from it. Similarly if we have attached new disk or changed its definition - with this flag it will instantly recreate instance with newly applied overrides.
If we omit this flag, the instance will continue to exist with no overrides changes applied until it gets restarted or recreated either manually or by autohealer or updater.

--remove-stateful-disks=DEVICE_NAME,[DEVICE_NAME,...]

List all device names which should be removed from current instance config.

--remove-stateful-metadata=KEY,[KEY,...]

List all stateful metadata keys which should be removed from current instance config.

--update-stateful-disk=[device-name=DEVICE-NAME],[mode=MODE],[source=SOURCE]

Disks considered stateful by the instance group. Usually, the managed instance group deletes disks when deleting instances; however, in the case of stateful disks, these disks are detached from the deleted instance and attached to new instances the managed instance group creates.
Besides preserving disks already attached to the instance by specifying only device names, user have an option to attach (and preserve) other existing persistent disk(s) to the given instance.
The same disk can be attached to many instances but only in read-only mode.
Use this argument multiple times to update multiple disks.
If stateful disk with given device-name exists in current instance config, its properties will be replaced by the newly provided ones. In other case new stateful disk definition will be added to the instance config.

device-name

Name under which disk is or will be attached.

source

Optional argument used to specify URI of existing persistent disk to attach under specified device-name.

mode

Specifies the mode of the disk to attach. Supported options are ro for read-only and rw for read-write. If omitted when source is specified, rw is used as a default.

--update-stateful-metadata=KEY=VALUE,[KEY=VALUE,...]

Additional metadata to be made available to the guest operating system on top of the metadata defined in the instance template.
Stateful metadata may be used to define a key/value pair specific for the one given instance to differentiate it from the other instances in the managed instance group.
Stateful metadata have priority over the metadata defined in the instance template. It means that stateful metadata defined for the keys already existing in the instance template override their values.
Each metadata entry is a key/value pair separated by an equals sign. Metadata keys must be unique and less than 128 bytes in length. Multiple entries can be passed to this flag, e.g., --update-stateful-metadata key-1=value-1,key-2=value-2,key-3=value-3.
If stateful metadata with the given key exists in current instance config, its value will be overridden with the newly provided one. If the key does not exist in the current instance config, a new key/value pair will be added.

At most one of these may be specified:

--region=REGION

Region of the managed instance group to update per instance config for. If not specified, you may be prompted to select a region.
A list of regions can be fetched by running:

$ gcloud compute regions list

Overrides the default compute/region property value for this command invocation.

--zone=ZONE

Zone of the managed instance group to update per instance config for. If not specified, you may be prompted to select a zone.
A list of zones can be fetched by running:

$ gcloud compute zones list

Overrides the default compute/zone property value for this command invocation.

GCLOUD WIDE FLAGS

NOTES