Description
When node auto-repair is disabled, persistent unhealthy nodes can require operator intervention, prolonging an outage. Auto-repair is enabled by default for new Standard node pools and is always enabled in Autopilot.
Potential impact
Prolonged node failures reduce availability and increase manual recovery work. Auto-repair does not resolve every failure immediately or replace data backups.
Remediation
Set management.auto_repair: true on Standard node pools. Maintain failure alerts and manual response procedures, and configure workloads to tolerate node replacement. Manage auto-upgrades for security patches separately.
Examples
These excerpts show node-pool management settings. cluster is a prepared cluster resource; supply the project and the path to a protected service-account JSON file separately.
Before
- name: create a node pool
google.cloud.gcp_container_node_pool:
name: my-pool
initial_node_count: 4
cluster: "{{ cluster }}"
location: us-central1-a
project: "{{ gcp_project }}"
auth_kind: serviceaccount
service_account_file: "{{ gcp_service_account_file }}"
state: present
management:
auto_repair: no
- name: create a node pool3
google.cloud.gcp_container_node_pool:
name: my-pool
initial_node_count: 4
cluster: "{{ cluster }}"
location: us-central1-a
project: "{{ gcp_project }}"
auth_kind: serviceaccount
service_account_file: "{{ gcp_service_account_file }}"
state: present
The first task disables auto-repair. The omission in the second does not mean auto-repair is disabled; check the actual node-pool state.
After
- name: create a node pool
google.cloud.gcp_container_node_pool:
name: my-pool
initial_node_count: 4
cluster: "{{ cluster }}"
location: us-central1-a
project: "{{ gcp_project }}"
auth_kind: serviceaccount
service_account_file: "{{ gcp_service_account_file }}"
state: present
management:
auto_repair: true
Auto-repair is explicitly enabled. Repairs follow the service’s health-check criteria.