A common pattern with GKE is to use a regional control plane, but nodes in a single zone. This makes most control plane operations a zero-downtime event and offers redundancy, but keeps actual workloads zonally colocated for latency reduction, inter-zone traffic reduction, and simplified storage classes (i.e., no regional disks or volume node affinity complexities). This is especially true for HPC/GPU/TPU node pools, but is also true for "system" components.

At present, this pattern is not possible with the GKE Cluster module system node pool, as the underlying Terraform module's node_locations setting, which override's the cluster's default zones, is not exposed.

This PR introduces a variable to optionally define specific system node pool zones. As this is an optional variable with null default, the module's default behavior and existing setups are unaffected.

Please take the following actions before submitting this pull request.

• ✅ Fork your PR branch from the Toolkit "develop" branch (not main)
• ✅ Test all changes with pre-commit in a local branch #
• ✅ Confirm that "make tests" passes all tests
• ✅ Add or modify unit tests to cover code changes
• ✅ Ensure that unit test coverage remains above 80%
• ✅ Update all applicable documentation
• ✅ Follow Cluster Toolkit Contribution guidelines #