gcloud compute instance-groups managed update - update a Compute Engine managed instance group
gcloud compute instance-groups managed update NAME [--description=DESCRIPTION] [--instance-redistribution-type=TYPE] [--list-managed-instances-results=MODE] [--remove-stateful-disks=DEVICE_NAME,[DEVICE_NAME,...]] [--stateful-disk=[auto-delete=AUTO-DELETE],[device-name=DEVICE-NAME]] [--target-distribution-shape=SHAPE] [--clear-autohealing | --initial-delay=INITIAL_DELAY --health-check=HEALTH_CHECK | --http-health-check=HTTP_HEALTH_CHECK | --https-health-check=HTTPS_HEALTH_CHECK] [--region=REGION | --zone=ZONE] [--update-policy-max-surge=MAX_SURGE --update-policy-max-unavailable=MAX_UNAVAILABLE --update-policy-minimal-action=UPDATE_POLICY_MINIMAL_ACTION --update-policy-most-disruptive-action=UPDATE_POLICY_MOST_DISRUPTIVE_ACTION --update-policy-replacement-method=UPDATE_POLICY_REPLACEMENT_METHOD --update-policy-type=UPDATE_TYPE] [GCLOUD_WIDE_FLAG ...]
Update a Compute Engine managed instance group.
gcloud compute instance-groups managed update allows you to specify or modify the description and group policies for an existing managed instance group, including the group's update policy and optional autohealing and stateful policies
The group's update policy defines how an updated VM configuration is applied to existing VMs in the group. For more information, see [Applying new configurations] https://cloud.google.com/compute/docs/instance-groups/updating-migs to VMs in a MIG.
A stateful policy defines which resources should be preserved across the group. When instances in the group are recreated, stateful resources are preserved. This command allows you to update stateful resources, specifically to add or remove stateful disks.
When updating the autohealing policy, you can specify the health check, initial delay, or both. If either field is unspecified, its value won't be modified. If --health-check is specified, the health check monitors the health of your application. Whenever the health check signal for an instance becomes UNHEALTHY, the autohealer recreates the instance.
If no health check exists, instance autohealing is triggered only by instance status: if an instance is not RUNNING, the group recreates it.
- NAME
Name of the managed instance group to update.
- --description=DESCRIPTION
An optional description for this group. To clear the description, set the value to an empty string.
- --instance-redistribution-type=TYPE
Specifies the type of the instance redistribution policy. An instance redistribution type lets you enable or disable automatic instance redistribution across zones to meet the group's target distribution shape.
An instance redistribution type can be specified only for a non-autoscaled regional managed instance group. By default it is set to proactive.
TYPE must be one of:
- none
The managed instance group does not redistribute instances across zones.
- proactive
The managed instance group proactively redistributes instances to meet its target distribution.
- --list-managed-instances-results=MODE
Pagination behavior for the group's listManagedInstances API method. This flag does not affect the group's gcloud or console list-instances behavior. By default it is set to pageless. MODE must be one of:
- pageless
Pagination is disabled for the group's listManagedInstances API method. maxResults and pageToken query parameters are ignored and all instances are returned in a single response.
- paginated
Pagination is enabled for the group's listManagedInstances API method. maxResults and pageToken query parameters are respected.
- --remove-stateful-disks=DEVICE_NAME,[DEVICE_NAME,...]
Remove stateful configuration for the specified disks.
- --stateful-disk=[auto-delete=AUTO-DELETE],[device-name=DEVICE-NAME]
Disks considered stateful by the instance group. Managed instance groups preserve and reattach stateful disks on VM autohealing, update, and recreate events.
Use this argument multiple times to update more disks.
If a stateful disk with the given device name already exists in the current instance configuration, its properties will be replaced by the newly provided ones. Otherwise, a new stateful disk definition will be added to the instance configuration.
- device-name
(Required) Device name of the disk to mark stateful.
- auto-delete
(Optional) Specifies the auto deletion policy of the stateful disk. The following options are available:
never: (Default) Never delete this disk. Instead, detach the disk when its instance is deleted.
on-permanent-instance-deletion: Delete the stateful disk when the instance that it's attached to is permanently deleted from the group; for example, when the instance is deleted manually or when the group size is decreased.
- --target-distribution-shape=SHAPE
Specifies how a regional managed instance group distributes its instances across zones within the region. The default shape is even. SHAPE must be one of:
- any
The group picks zones for creating VM instances to fulfill the requested number of VMs within present resource constraints and to maximize utilization of unused zonal reservations. Recommended for batch workloads that do not require high availability.
- balanced
The group prioritizes acquisition of resources, scheduling VMs in zones where resources are available while distributing VMs as evenly as possible across selected zones to minimize the impact of zonal failure. Recommended for highly available serving or batch workloads that do not require autoscaling.
- even
The group schedules VM instance creation and deletion to achieve and maintain an even number of managed instances across the selected zones. The distribution is even when the number of managed instances does not differ by more than 1 between any two zones. Recommended for highly available serving workloads.
- At most one of these can be specified:
- --clear-autohealing
Clears all autohealing policy fields for the managed instance group.
- --initial-delay=INITIAL_DELAY
Specifies the length of time during which the instance is known to be initializing and should not be autohealed even if unhealthy. This value cannot be greater than 1 hour. See $ gcloud topic datetimes for information on duration formats.
- At most one of these can be specified:
- --health-check=HEALTH_CHECK
Name of the health check to operate on.
- --http-health-check=HTTP_HEALTH_CHECK
(DEPRECATED) HTTP health check object used for autohealing instances in this group.
HttpHealthCheck is deprecated. Use --health-check instead.
- --https-health-check=HTTPS_HEALTH_CHECK
(DEPRECATED) HTTPS health check object used for autohealing instances in this group.
HttpsHealthCheck is deprecated. Use --health-check instead.
- At most one of these can be specified:
- --region=REGION
Region of the managed instance group to update. If not specified, you might be prompted to select a region (interactive mode only).
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. If not specified, you might be prompted to select a zone (interactive mode only).
A list of zones can be fetched by running:
$ gcloud compute zones list
Overrides the default compute/zone property value for this command invocation.
- Parameters for setting update policy for this managed instance group.
- --update-policy-max-surge=MAX_SURGE
Maximum additional number of VMs that can be created during the update process. This can be a fixed number (e.g. 5) or a percentage of size to the managed instance group (e.g. 10%). Defaults to 0 if the managed instance group has stateful configuration, or to the number of zones in which it operates otherwise.
- --update-policy-max-unavailable=MAX_UNAVAILABLE
Maximum number of VMs that can be unavailable during the update process. This can be a fixed number (e.g. 5) or a percentage of size to the managed instance group (e.g. 10%). Defaults to the number of zones in which the managed instance group operates.
- --update-policy-minimal-action=UPDATE_POLICY_MINIMAL_ACTION
Use this flag to minimize disruption as much as possible or to apply a more disruptive action than is strictly necessary. The MIG performs at least this action on each VM while updating. If the update requires a more disruptive action than the one specified here, then the more disruptive action is performed. UPDATE_POLICY_MINIMAL_ACTION must be one of:
- none
No action
- refresh
Apply the new configuration without stopping VMs, if possible. For example, use ``refresh`` to apply changes that only affect metadata or additional disks.
- restart
Apply the new configuration without replacing VMs, if possible. For example, stopping VMs and starting them again is sufficient to apply changes to machine type.
- replace
Replace old VMs according to the --update-policy-replacement-method flag.
- --update-policy-most-disruptive-action=UPDATE_POLICY_MOST_DISRUPTIVE_ACTION
Use this flag to prevent an update if it requires more disruption than you can afford. At most, the MIG performs the specified action on each VM while updating. If the update requires a more disruptive action than the one specified here, then the update fails and no changes are made. UPDATE_POLICY_MOST_DISRUPTIVE_ACTION must be one of:
- none
No action
- refresh
Apply the new configuration without stopping VMs, if possible. For example, use ``refresh`` to apply changes that only affect metadata or additional disks.
- restart
Apply the new configuration without replacing VMs, if possible. For example, stopping VMs and starting them again is sufficient to apply changes to machine type.
- replace
Replace old VMs according to the --update-policy-replacement-method flag.
- --update-policy-replacement-method=UPDATE_POLICY_REPLACEMENT_METHOD
Type of replacement method. Specifies what action will be taken to update VMs. Defaults to ``recreate`` if the managed instance group has stateful configuration, or to ``substitute`` otherwise. UPDATE_POLICY_REPLACEMENT_METHOD must be one of:
- recreate
Recreate VMs and preserve the VM names. The VM IDs and creation timestamps might change.
- substitute
Delete old VMs and create VMs with new names.
- --update-policy-type=UPDATE_TYPE
Specifies the type of update process. You can specify either ``proactive`` so that the managed instance group proactively executes actions in order to bring VMs to their target versions or ``opportunistic`` so that no action is proactively executed but the update will be performed as part of other actions. UPDATE_TYPE must be one of:
- opportunistic
Do not proactively replace VMs. Create new VMs and delete old ones on resizes of the group and when you target specific VMs to be updated or recreated.
- proactive
Replace VMs proactively.
These flags are available to all commands: --access-token-file, --account, --billing-project, --configuration, --flags-file, --flatten, --format, --help, --impersonate-service-account, --log-http, --project, --quiet, --trace-token, --user-output-enabled, --verbosity.
Run $ gcloud help for details.
These variants are also available:
$ gcloud alpha compute instance-groups managed update
$ gcloud beta compute instance-groups managed update