Release 0.14.1

goldens/Basic_cluster_create.txt

@@ -81,16 +81,15 @@ kubectl apply --server-side --force-conflicts -f https://github.com/kubernetes-s
 [XPK] Task: `Wait for Kueue to be available` is implemented by the following command not running since it is a dry run. 
 kubectl wait deploy/kueue-controller-manager -nkueue-system --for=condition=available --timeout=10m
 [XPK] Applying following Kueue resources:
----
 apiVersion: kueue.x-k8s.io/v1beta1
 kind: ResourceFlavor
 metadata:
   name: "1xtpu7x-8"
 spec:
   nodeLabels: {"cloud.google.com/gke-tpu-accelerator": "tpu7x", "cloud.google.com/gke-tpu-topology": "2x2x1"}
 
-
 ---
+
 apiVersion: kueue.x-k8s.io/v1beta1
 kind: AdmissionCheck
 metadata:
@@ -175,7 +174,7 @@ value: 1000
 globalDefault: false
 description: "Very High"
 [XPK] Task: `Applying Kueue Custom Resources` is implemented by the following command not running since it is a dry run. 
-kubectl apply -f a1fe8e014a200d6489b8871301a9e80de7e6f45e94b61ad0e60f40f254711bec
+kubectl apply -f ce52d2868b681f478f3f12e5696b1609e68b442a32f7f82603ba7064b825cf4f
 [XPK] Task: `Count total nodes` is implemented by the following command not running since it is a dry run. 
 kubectl get node --no-headers | wc -l
 [XPK] Try 1: Updating Kueue Controller Manager resources

goldens/Cluster_create_private.txt

@@ -86,25 +86,24 @@ kubectl apply --server-side --force-conflicts -f https://github.com/kubernetes-s
 [XPK] Task: `Wait for Kueue to be available` is implemented by the following command not running since it is a dry run. 
 kubectl wait deploy/kueue-controller-manager -nkueue-system --for=condition=available --timeout=10m
 [XPK] Applying following Kueue resources:
----
 apiVersion: kueue.x-k8s.io/v1beta1
 kind: ResourceFlavor
 metadata:
   name: "1xv5p-8"
 spec:
   nodeLabels: {"cloud.google.com/gke-tpu-accelerator": "tpu-v5p-slice", "cloud.google.com/gke-tpu-topology": "2x2x1"}
 
-
 ---
+
 apiVersion: kueue.x-k8s.io/v1beta1
 kind: ResourceFlavor
 metadata:
   name: "cpu-user"
 spec:
   nodeLabels: {"cloud.google.com/gke-nodepool": "cpu-np"}
 
-
 ---
+
 apiVersion: kueue.x-k8s.io/v1beta1
 kind: AdmissionCheck
 metadata:
@@ -189,7 +188,7 @@ value: 1000
 globalDefault: false
 description: "Very High"
 [XPK] Task: `Applying Kueue Custom Resources` is implemented by the following command not running since it is a dry run. 
-kubectl apply -f 0ff2bce892606d1497f21fc7b2cea78a4ee103094ce0f509211f3f9730536ad6
+kubectl apply -f 1838a525f73d2cfd19aa616665e8b33fcc4ea10ba8d6015a9307109a6be6d372
 [XPK] Task: `Count total nodes` is implemented by the following command not running since it is a dry run. 
 kubectl get node --no-headers | wc -l
 [XPK] Try 1: Updating Kueue Controller Manager resources

goldens/Cluster_create_with_gb200-4.txt

@@ -85,16 +85,15 @@ kubectl apply --server-side --force-conflicts -f https://github.com/kubernetes-s
 [XPK] Task: `Wait for Kueue to be available` is implemented by the following command not running since it is a dry run. 
 kubectl wait deploy/kueue-controller-manager -nkueue-system --for=condition=available --timeout=10m
 [XPK] Applying following Kueue resources:
----
 apiVersion: kueue.x-k8s.io/v1beta1
 kind: ResourceFlavor
 metadata:
   name: "1xgb200-4"
 spec:
   nodeLabels: {"cloud.google.com/gke-accelerator": "nvidia-gb200"}
 
-
 ---
+
 apiVersion: kueue.x-k8s.io/v1beta1
 kind: AdmissionCheck
 metadata:
@@ -179,7 +178,7 @@ value: 1000
 globalDefault: false
 description: "Very High"
 [XPK] Task: `Applying Kueue Custom Resources` is implemented by the following command not running since it is a dry run. 
-kubectl apply -f f807069b73747a423ec0d1915b2e919cfde400b01654de15746b566709b80f7e
+kubectl apply -f 5c5d70f8d2bbedea9acccd9c1a153e2f55efd31cc61d2b55ecdd4a8f009fab11
 [XPK] Task: `Count total nodes` is implemented by the following command not running since it is a dry run. 
 kubectl get node --no-headers | wc -l
 [XPK] Try 1: Updating Kueue Controller Manager resources

goldens/NAP_cluster-create.txt

@@ -92,16 +92,15 @@ kubectl apply --server-side --force-conflicts -f https://github.com/kubernetes-s
 [XPK] Task: `Wait for Kueue to be available` is implemented by the following command not running since it is a dry run. 
 kubectl wait deploy/kueue-controller-manager -nkueue-system --for=condition=available --timeout=10m
 [XPK] Applying following Kueue resources:
----
 apiVersion: kueue.x-k8s.io/v1beta1
 kind: ResourceFlavor
 metadata:
   name: "1xtpu7x-8"
 spec:
   nodeLabels: {"cloud.google.com/gke-tpu-accelerator": "tpu7x"}
 
-
 ---
+
 apiVersion: kueue.x-k8s.io/v1beta1
 kind: AdmissionCheck
 metadata:
@@ -186,7 +185,7 @@ value: 1000
 globalDefault: false
 description: "Very High"
 [XPK] Task: `Applying Kueue Custom Resources` is implemented by the following command not running since it is a dry run. 
-kubectl apply -f e5fccf0957dcb7f60400bb4e28ce8c5fc251a9aeb6d67793dc119554d13dc900
+kubectl apply -f 40a7aac5b047c750ee98477984af3d46acd60d164d852eccd1b47a21c4155f2d
 [XPK] Task: `Count total nodes` is implemented by the following command not running since it is a dry run. 
 kubectl get node --no-headers | wc -l
 [XPK] Try 1: Updating Kueue Controller Manager resources

goldens/NAP_cluster-create_with_pathways.txt

@@ -93,25 +93,24 @@ kubectl apply --server-side --force-conflicts -f https://github.com/kubernetes-s
 [XPK] Task: `Wait for Kueue to be available` is implemented by the following command not running since it is a dry run. 
 kubectl wait deploy/kueue-controller-manager -nkueue-system --for=condition=available --timeout=10m
 [XPK] Applying following Kueue resources:
----
 apiVersion: kueue.x-k8s.io/v1beta1
 kind: ResourceFlavor
 metadata:
   name: "1xtpu7x-8"
 spec:
   nodeLabels: {"cloud.google.com/gke-tpu-accelerator": "tpu7x"}
 
-
 ---
+
 apiVersion: kueue.x-k8s.io/v1beta1
 kind: ResourceFlavor
 metadata:
   name: "cpu-user"
 spec:
   nodeLabels: {"cloud.google.com/gke-nodepool": "cpu-np"}
 
-
 ---
+
 apiVersion: kueue.x-k8s.io/v1beta1
 kind: AdmissionCheck
 metadata:
@@ -196,7 +195,7 @@ value: 1000
 globalDefault: false
 description: "Very High"
 [XPK] Task: `Applying Kueue Custom Resources` is implemented by the following command not running since it is a dry run. 
-kubectl apply -f edda4d2ffefedaabd6f77eb6d07a05c1bac829ed90a4c3983b21ded280136557
+kubectl apply -f f89effb1f55aef327018037d75f743b5c62d59f1f62fddadaaa31f72e5e07bdf
 [XPK] Task: `Count total nodes` is implemented by the following command not running since it is a dry run. 
 kubectl get node --no-headers | wc -l
 [XPK] Try 1: Updating Kueue Controller Manager resources

goldens/Workload_create.txt

@@ -2,7 +2,6 @@ $ python3 xpk.py workload create --project=golden-project --zone=us-central1-a -
 [XPK] Starting xpk
 [XPK] Task: `Check if Workload Already Exists` is implemented by the following command not running since it is a dry run. 
 kubectl get workloads -o=custom-columns='Jobset:.metadata.ownerReferences[0].name'
-[XPK] Starting workload create
 [XPK] Task: `GKE Cluster Get ConfigMap` is implemented by the following command not running since it is a dry run. 
 kubectl get configmap golden-cluster-resources-configmap -o=custom-columns="ConfigData:data" --no-headers=true
 [XPK] Starting workload create
@@ -25,7 +24,7 @@ docker tag dry-run-runner gcr.io/golden-project/dry-run-runner:prefix-current
 [XPK] Task: `Upload Docker Image` is implemented by the following command not running since it is a dry run. 
 docker push gcr.io/golden-project/dry-run-runner:prefix-current
 [XPK] Task: `Creating Workload` is implemented by the following command not running since it is a dry run. 
-kubectl apply -f 492d3bb4e1055d9d47679ed9c3ba617c304f47bac9b83fea3c14507b04a65453
+kubectl apply -f abc33690f7a11b2ba50a8f949970fd3ba812f088367b7f64260729f01f41a231
 [XPK] Task: `GKE Dashboard List` is implemented by the following command not running since it is a dry run. 
 gcloud monitoring dashboards list --project=golden-project --filter="displayName:'GKE - TPU Monitoring Dashboard'" --format="value(name)" --verbosity=error
 [XPK] Check statistics and outlier mode of GKE metrics here: https://console.cloud.google.com/monitoring/dashboards/builder/0?project=golden-project&f.rlabel.cluster_name.ClusterName=golden-cluster. To view the metric data for your workload, select golden-workload from the JobName filter on the dashboard.

goldens/Workload_create_pathways.txt

@@ -2,7 +2,6 @@ $ python3 xpk.py workload create-pathways --project=golden-project --zone=us-cen
 [XPK] Starting xpk
 [XPK] Task: `Check if Workload Already Exists` is implemented by the following command not running since it is a dry run. 
 kubectl get workloads -o=custom-columns='Jobset:.metadata.ownerReferences[0].name'
-[XPK] Starting workload create
 [XPK] Task: `GKE Cluster Get ConfigMap` is implemented by the following command not running since it is a dry run. 
 kubectl get configmap golden-cluster-resources-configmap -o=custom-columns="ConfigData:data" --no-headers=true
 [XPK] Starting workload create

pyproject.toml

@@ -75,9 +75,12 @@ dev = [
 version = {attr = "xpk.core.config.__version__"}
 
 [tool.setuptools]
-packages = ["xpk", "xpk.parser", "xpk.core", "xpk.commands", "xpk.api", "xpk.templates", "xpk.utils", "xpk.core.blueprint", "xpk.core.remote_state", "xpk.core.workload_decorators"]
 package-dir = {"" = "src"}
-package-data = {"xpk.api" = ["storage_crd.yaml"], "xpk.templates" = ["storage.yaml"]}
+packages = { find = { where = ["src"] } }
+
+[tool.setuptools.package-data]
+"xpk" = ["templates/*"]
+"xpk.api" = ["*.yaml"]
 
 [tool.pyink]
 # Formatting configuration to follow Google style-guide.

src/xpk/commands/cluster.py

@@ -16,6 +16,7 @@
 
 from tabulate import tabulate
 
+from ..utils.feature_flags import FeatureFlags
 from ..core.capacity import H100_DEVICE_TYPE, H200_DEVICE_TYPE, B200_DEVICE_TYPE
 from ..core.cluster import (
     get_all_clusters_programmatic,
@@ -75,9 +76,9 @@
 from ..utils.execution_context import is_dry_run
 from ..utils.validation import validate_dependencies_list, SystemDependency, should_validate_dependencies
 from . import cluster_gcluster
-from .common import set_cluster_command
+from .common import set_cluster_command, validate_sub_slicing_system
 from jinja2 import Environment, FileSystemLoader
-from ..utils.templates import TEMPLATE_PATH
+from ..utils.templates import get_templates_absolute_path
 import shutil
 import os
 
@@ -200,6 +201,11 @@ def cluster_adapt(args) -> None:
   xpk_exit(0)
 
 
+def _validate_cluster_create_args(args, system: SystemCharacteristics):
+  if FeatureFlags.SUB_SLICING_ENABLED and args.sub_slicing:
+    validate_sub_slicing_system(system)
+
+
 def cluster_create(args) -> None:
   """Function around cluster creation.
 
@@ -212,12 +218,14 @@ def cluster_create(args) -> None:
         SystemDependency.KJOB,
         SystemDependency.GCLOUD,
     ])
-  system, return_code = get_system_characteristics(args)
 
+  system, return_code = get_system_characteristics(args)
   if return_code > 0 or system is None:
     xpk_print('Fetching system characteristics failed!')
     xpk_exit(return_code)
 
+  _validate_cluster_create_args(args, system)
+
   xpk_print(f'Starting cluster create for cluster {args.cluster}:', flush=True)
   add_zone_and_project(args)
 
@@ -426,7 +434,9 @@ def cluster_cacheimage(args) -> None:
       system.accelerator_type
   ].accelerator_label
 
-  template_env = Environment(loader=FileSystemLoader(TEMPLATE_PATH))
+  template_env = Environment(
+      loader=FileSystemLoader(searchpath=get_templates_absolute_path())
+  )
   cluster_preheat_yaml = template_env.get_template(CLUSTER_PREHEAT_JINJA_FILE)
   rendered_yaml = cluster_preheat_yaml.render(
       cachekey=args.cache_key,
@@ -1251,6 +1261,9 @@ def install_kueue(args, system: SystemCharacteristics, autoprovisioning_config):
           memory_limit=args.memory_limit,
           cpu_limit=args.cpu_limit,
           is_pathways_cluster=args.enable_pathways,
+          configure_sub_slicing=(
+              FeatureFlags.SUB_SLICING_ENABLED and args.sub_slicing
+          ),
       ),
   )
 

src/xpk/commands/cluster_gcluster.py

@@ -16,6 +16,7 @@
 
 import os
 
+from ..utils.feature_flags import FeatureFlags
 from ..utils.execution_context import is_dry_run
 from ..core.kueue_manager import KueueConfig, KueueManager
 from ..core.nap import enable_autoprovisioning_on_cluster
@@ -159,6 +160,9 @@ def __install_kueue(args) -> int:
           cpu_limit=args.cpu_limit,
           is_pathways_cluster=args.enable_pathways,
           flex=args.flex,
+          configure_sub_slicing=(
+              FeatureFlags.SUB_SLICING_ENABLED and args.sub_slicing
+          ),
       ),
       tolerations=tolerations,
   )

src/xpk/commands/cluster_test.py

@@ -0,0 +1,92 @@
+"""
+Copyright 2025 Google LLC
+
+Licensed under the Apache License, Version 2.0 (the "License");
+you may not use this file except in compliance with the License.
+You may obtain a copy of the License at
+
+     https://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software
+distributed under the License is distributed on an "AS IS" BASIS,
+WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+See the License for the specific language governing permissions and
+limitations under the License.
+"""
+
+from argparse import Namespace
+from dataclasses import dataclass
+from unittest.mock import MagicMock
+import pytest
+
+from xpk.commands.cluster import _validate_cluster_create_args
+from xpk.core.system_characteristics import SystemCharacteristics, UserFacingNameToSystemCharacteristics
+from xpk.utils.feature_flags import FeatureFlags
+
+
+@dataclass
+class _Mocks:
+  common_print_mock: MagicMock
+  common_exit_mock: MagicMock
+
+
+@pytest.fixture
+def mock_common_print_and_exit(mocker):
+  common_print_mock = mocker.patch(
+      'xpk.commands.common.xpk_print',
+      return_value=None,
+  )
+  common_exit_mock = mocker.patch(
+      'xpk.commands.common.xpk_exit',
+      return_value=None,
+  )
+  return _Mocks(
+      common_print_mock=common_print_mock, common_exit_mock=common_exit_mock
+  )
+
+
+DEFAULT_TEST_SYSTEM: SystemCharacteristics = (
+    UserFacingNameToSystemCharacteristics['l4-1']
+)
+SUB_SLICING_SYSTEM: SystemCharacteristics = (
+    UserFacingNameToSystemCharacteristics['v6e-4x4']
+)
+
+
+def test_validate_cluster_create_args_for_correct_args_pass(
+    mock_common_print_and_exit: _Mocks,
+):
+  args = Namespace()
+
+  _validate_cluster_create_args(args, DEFAULT_TEST_SYSTEM)
+
+  assert mock_common_print_and_exit.common_print_mock.call_count == 0
+  assert mock_common_print_and_exit.common_exit_mock.call_count == 0
+
+
+def test_validate_cluster_create_args_for_correct_sub_slicing_args_pass(
+    mock_common_print_and_exit: _Mocks,
+):
+  FeatureFlags.SUB_SLICING_ENABLED = True
+  args = Namespace(sub_slicing=True)
+
+  _validate_cluster_create_args(args, SUB_SLICING_SYSTEM)
+
+  assert mock_common_print_and_exit.common_print_mock.call_count == 0
+  assert mock_common_print_and_exit.common_exit_mock.call_count == 0
+
+
+def test_validate_cluster_create_args_for_not_supported_system_throws(
+    mock_common_print_and_exit: _Mocks,
+):
+  FeatureFlags.SUB_SLICING_ENABLED = True
+  args = Namespace(sub_slicing=True)
+
+  _validate_cluster_create_args(args, DEFAULT_TEST_SYSTEM)
+
+  assert mock_common_print_and_exit.common_print_mock.call_count == 1
+  assert (
+      mock_common_print_and_exit.common_print_mock.call_args[0][0]
+      == 'Error: l4-1 does not support Sub-slicing.'
+  )
+  assert mock_common_print_and_exit.common_exit_mock.call_count == 1

src/xpk/commands/common.py

@@ -67,3 +67,9 @@ def is_TAS_possible(
       system_characteristics.device_type != H100_MEGA_DEVICE_TYPE
       or capacity_type == CapacityType.RESERVATION
   )
+
+
+def validate_sub_slicing_system(system: SystemCharacteristics):
+  if not system.supports_sub_slicing:
+    xpk_print(f'Error: {system.device_type} does not support Sub-slicing.')
+    xpk_exit(1)

src/xpk/commands/kind.py

@@ -110,6 +110,7 @@ def cluster_create(args) -> None:
           cpu_limit=0,
           is_pathways_cluster=False,
           flex=False,
+          configure_sub_slicing=False,
       ),
   )