xpk 1.0.0__py3-none-any.whl → 1.1.0__py3-none-any.whl

xpk/commands/inspector_test.py

@@ -0,0 +1,142 @@
+"""
+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.
+"""
+
+import pytest
+from unittest import mock
+from xpk.commands import inspector
+from xpk.core.testing.commands_tester import CommandsTester
+
+
+@pytest.fixture
+def args():
+  args = mock.Mock()
+  args.print_to_terminal = False
+  return args
+
+
+@pytest.fixture
+def commands_tester(mocker):
+  return CommandsTester(
+      mocker,
+      run_command_for_value_path="xpk.commands.inspector.run_command_for_value",
+  )
+
+
+@pytest.fixture
+def mock_has_super_slicing_enabled(mocker):
+  return mocker.patch("xpk.commands.inspector.has_super_slicing_enabled")
+
+
+@pytest.fixture
+def mock_append_tmp_file(mocker):
+  return mocker.patch("xpk.commands.inspector.append_tmp_file")
+
+
+@pytest.fixture
+def mock_xpk_print(mocker):
+  return mocker.patch("xpk.commands.inspector.xpk_print")
+
+
+def test_inspector_run_slice_controller_helper_no_super_slicing(
+    args: mock.Mock,
+    commands_tester: CommandsTester,
+    mock_has_super_slicing_enabled: mock.Mock,
+    mock_append_tmp_file: mock.Mock,
+):
+  mock_has_super_slicing_enabled.return_value = (0, False)
+
+  inspector.inspector_run_slice_controller_helper(args, "test_file")
+  commands_tester.assert_command_not_run(
+      "kubectl logs deployment slice-controller-controller-manager"
+  )
+  commands_tester.assert_command_not_run(
+      "kubectl describe deployment slice-controller-controller-manager"
+  )
+  mock_append_tmp_file.assert_not_called()
+
+
+def test_inspector_run_slice_controller_helper_with_super_slicing_success(
+    args: mock.Mock,
+    commands_tester: CommandsTester,
+    mock_has_super_slicing_enabled: mock.Mock,
+    mock_append_tmp_file: mock.Mock,
+):
+  commands_tester.set_result_for_command(
+      (0, "some logs"),
+      "kubectl",
+      "logs",
+      "deployment slice-controller-controller-manager",
+  )
+  commands_tester.set_result_for_command(
+      (0, "some details"),
+      "kubectl",
+      "describe",
+      "deployment slice-controller-controller-manager",
+  )
+  mock_has_super_slicing_enabled.return_value = (0, True)
+
+  inspector.inspector_run_slice_controller_helper(args, "test_file")
+
+  commands_tester.assert_command_run(
+      "kubectl logs deployment slice-controller-controller-manager"
+  )
+  commands_tester.assert_command_run(
+      "kubectl describe deployment slice-controller-controller-manager"
+  )
+
+  mock_append_tmp_file.assert_called()
+  call_args_list = mock_append_tmp_file.call_args_list
+  assert any(
+      "Super-slicing topology set up" in args[0] for args, _ in call_args_list
+  )
+  assert any("some logs" in args[0] for args, _ in call_args_list)
+  assert any("some details" in args[0] for args, _ in call_args_list)
+
+
+def test_inspector_run_slice_controller_helper_with_slice_controller_not_found(
+    args: mock.Mock,
+    commands_tester: CommandsTester,
+    mock_has_super_slicing_enabled: mock.Mock,
+    mock_append_tmp_file: mock.Mock,
+    mock_xpk_print: mock.Mock,
+):
+  commands_tester.set_result_for_command(
+      (1, "Error: Deployment not found"),
+      "kubectl",
+      "deployment slice-controller-controller-manager",
+  )
+  mock_has_super_slicing_enabled.return_value = (0, True)
+
+  inspector.inspector_run_slice_controller_helper(args, "test_file")
+
+  commands_tester.assert_command_run(
+      "kubectl describe deployment slice-controller-controller-manager"
+  )
+  commands_tester.assert_command_run(
+      "kubectl logs deployment slice-controller-controller-manager"
+  )
+
+  mock_append_tmp_file.assert_called()
+  call_args_list = mock_append_tmp_file.call_args_list
+  assert any(
+      "Super-slicing topology set up" in args[0] for args, _ in call_args_list
+  )
+
+  mock_xpk_print.assert_called()
+  call_args_list = mock_xpk_print.call_args_list
+  assert any(
+      "Error: Deployment not found" in args[0] for args, _ in call_args_list
+  )

xpk/commands/workload.py

@@ -54,6 +54,7 @@ from ..core.resources import get_cluster_capacity_type, get_cluster_system_chara
 from ..core.resources import ConfigMapType, get_cluster_configmap
 from ..core.nodepool import ensure_resource_policy_exists
 from ..core.scheduling import (
+    ONE_TO_ONE_REPLICA_NODE_POOL_ASSIGNMENT_ANNOTATION,
     WorkloadScheduling,
     check_if_workload_can_schedule,
     create_tpu_machine_type,
@@ -99,7 +100,7 @@ from ..utils.file import write_tmp_file
 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 is_TAS_possible
+from .common import is_GPU_TAS_possible
 from jinja2 import Environment, FileSystemLoader
 from ..utils.templates import get_templates_absolute_path
 
@@ -111,7 +112,7 @@ metadata:
     kueue.x-k8s.io/queue-name: {local_queue_name}  # Name of the LocalQueue
     xpk.google.com/workload: {args.workload}
   annotations:
-    alpha.jobset.sigs.k8s.io/exclusive-topology: cloud.google.com/gke-nodepool # 1:1 job replica to node pool assignment
+    {jobset_annotations}
 spec:
   ttlSecondsAfterFinished: {args.ttl_seconds_after_finished}
   failurePolicy:
@@ -490,13 +491,21 @@ def workload_create(args) -> None:
         - PodFailurePolicy"""
     restart_on_exit_codes_list = get_restart_exit_codes(args)
     restart_on_exit_codes = ','.join(map(str, restart_on_exit_codes_list))
-    pod_failure_policy = f"""
+
+    pod_failure_policy = """
           podFailurePolicy:
             rules:
+          """
+    docker_image = get_main_container_docker_image(args, workload_system)
+    for i in range(workload_system.parallel_containers):
+      docker_image_sufix = (
+          f'-{i + 1}' if workload_system.parallel_containers > 1 else ''
+      )
+      pod_failure_policy += f"""
             - action: FailJob
               onPodConditions: []
               onExitCodes:
-                containerName: {get_main_container_docker_image(args, workload_system)}
+                containerName: {docker_image}{docker_image_sufix}
                 operator: NotIn
                 values: [{restart_on_exit_codes}]"""
 
@@ -534,11 +543,10 @@ def workload_create(args) -> None:
     capacity_type = get_cluster_capacity_type(args)
 
     annotations = (
-        (
-            'kueue.x-k8s.io/podset-preferred-topology:'
-            ' "cloud.google.com/gce-topology-host"'
+        'kueue.x-k8s.io/podset-preferred-topology: "kubernetes.io/hostname"'
+        if is_GPU_TAS_possible(
+            cluster_system, capacity_type, args.cluster, args.zone, args.project
         )
-        if is_TAS_possible(cluster_system, capacity_type)
         else ''
     )
 
@@ -648,9 +656,15 @@ def workload_create(args) -> None:
         if use_super_slicing
         else ''
     )
+    jobset_annotations = (
+        ''
+        if use_super_slicing or use_sub_slicing
+        else ONE_TO_ONE_REPLICA_NODE_POOL_ASSIGNMENT_ANNOTATION
+    )
 
     yml_string = WORKLOAD_CREATE_YAML.format(
         args=args,
+        jobset_annotations=jobset_annotations,
         container=container,
         vms_per_slice=workload_system.vms_per_slice,
         affinity=get_cpu_affinity(workload_system.accelerator_type),

xpk/commands/workload_test.py

@@ -23,6 +23,7 @@ from ..core.scheduling import WorkloadScheduling
 from ..core.system_characteristics import DockerPlatform, SystemCharacteristics, AcceleratorType, UserFacingNameToSystemCharacteristics, GpuConfig
 from .workload import workload_create
 from .cluster_test import construct_args
+from ..core.docker_container import get_user_workload_container as real_get_user_workload_container
 
 
 SYSTEM_CHARACTERISTICS = SystemCharacteristics(
@@ -58,7 +59,7 @@ class _WorkloadCreateMocks:
   validate_dependencies_list: MagicMock
   write_tmp_file: MagicMock
   get_cluster_capacity_type: MagicMock
-  is_TAS_possible: MagicMock
+  is_GPU_TAS_possible: MagicMock
   get_cluster_location: MagicMock
   xpk_exit: MagicMock
   run_command_with_updates: MagicMock
@@ -113,8 +114,8 @@ def workload_create_mocks(mocker) -> _WorkloadCreateMocks:
           'xpk.commands.workload.get_cluster_capacity_type',
           return_value='on-demand',
       ),
-      is_TAS_possible=mocker.patch(
-          'xpk.commands.workload.is_TAS_possible', return_value=False
+      is_GPU_TAS_possible=mocker.patch(
+          'xpk.commands.workload.is_GPU_TAS_possible', return_value=False
       ),
       get_cluster_location=mocker.patch(
           'xpk.commands.workload.get_cluster_location',
@@ -206,3 +207,69 @@ def test_workload_create_dry_run_with_output_file(mocker):
   written_content = mock_open.return_value.write.call_args[0][0]
   assert 'test-workload' in written_content
   assert 'cloud.google.com/gke-tpu-topology: 8x8' in written_content
+
+
+def test_workload_create_multi_container_for_tpu7x(
+    workload_create_mocks: _WorkloadCreateMocks,
+    mocker,
+):
+  """Tests that the generated YAML for a multi-container workload has correct pod failure policy and container structure."""
+
+  # Enable dry_run to prevent external calls like get_storages_to_mount -> gcloud
+  mocker.patch('xpk.utils.execution_context.dry_run', True)
+
+  # Mock dependencies required by get_user_workload_container -> get_main_container
+  mocker.patch(
+      'xpk.core.docker_container.setup_docker_image',
+      return_value=(0, 'dummy-image'),
+  )
+  mocker.patch(
+      'xpk.core.docker_container.get_gke_debugging_dashboard', return_value=None
+  )
+
+  # Use the real get_user_workload_container to test integration
+  workload_create_mocks.get_user_workload_container.side_effect = (
+      real_get_user_workload_container
+  )
+
+  args = construct_args(
+      workload='test-workload',
+      command='echo hello',
+      num_nodes=1,
+      tpu_type='tpu7x-2x2x2',
+      restart_on_exit_codes=None,
+      docker_name='test-docker',
+      deploy_stacktrace_sidecar=False,
+      enable_debug_logs=False,
+      scheduler='default-scheduler',
+  )
+  workload_create(args)
+
+  assert workload_create_mocks.write_tmp_file.called
+  yaml_content = workload_create_mocks.write_tmp_file.call_args[0][0]
+  jobset = yaml.safe_load(yaml_content)
+
+  # Verify Pod Failure Policy
+  pod_failure_rules = jobset['spec']['replicatedJobs'][0]['template']['spec'][
+      'podFailurePolicy'
+  ]['rules']
+  # Should have 2 rules for multi_container
+  assert len(pod_failure_rules) == 2
+  assert pod_failure_rules[0]['onExitCodes']['containerName'].endswith('-1')
+  assert pod_failure_rules[1]['onExitCodes']['containerName'].endswith('-2')
+
+  # Verify Containers
+  # Navigate to the containers list in the YAML
+  containers = jobset['spec']['replicatedJobs'][0]['template']['spec'][
+      'template'
+  ]['spec']['containers']
+
+  assert len(containers) == 2
+  assert containers[0]['name'].endswith('-1')
+  assert containers[1]['name'].endswith('-2')
+  assert containers[0]['image'] == 'dummy-image'
+  assert containers[1]['image'] == 'dummy-image'
+
+  # Check if resources are split correctly (4 chips / 2 containers = 2 chips)
+  assert containers[0]['resources']['limits']['google.com/tpu'] == 2
+  assert containers[1]['resources']['limits']['google.com/tpu'] == 2

xpk/core/blueprint/blueprint_generator.py

@@ -24,6 +24,7 @@ from packaging.version import parse
 from ...utils.console import xpk_exit, xpk_print
 from ...utils.versions import ReleaseChannel
 from ...utils.file import ensure_directory_exists
+from ...utils.templates import get_templates_absolute_path
 
 
 from ..capacity import (
@@ -51,9 +52,9 @@ supported_device_types = {
     a4_device_type,
 }
 blueprint_dependencies_dir = {
-    a3mega_device_type: "src/xpk/blueprints/a3mega",
-    a3ultra_device_type: "src/xpk/blueprints/a3ultra",
-    a4_device_type: "src/xpk/blueprints/a4",
+    a3mega_device_type: get_templates_absolute_path("blueprints/a3mega"),
+    a3ultra_device_type: get_templates_absolute_path("blueprints/a3ultra"),
+    a4_device_type: get_templates_absolute_path("blueprints/a4"),
 }
 
 cluster_toolkit_url = "github.com/GoogleCloudPlatform/cluster-toolkit"
@@ -63,7 +64,7 @@ common_cluster_labels = {"gke_product_type": "xpk"}
 
 class BlueprintGeneratorOutput:
   """BlueprintGeneratorOutput is a class containing fields with output blueprint file path and path to blueprint dependencies.
-  Atributes:
+  Attributes:
   - blueprint_file (str) : path to generated blueprint file.
   - blueprint_dependencies (str) : path to directory containing blueprint dependencies.
   """
@@ -75,7 +76,7 @@ class BlueprintGeneratorOutput:
 
 class BlueprintGenerator:
   """BlueprintGenerator is a class for generating blueprints
-  Atributes:
+  Attributes:
   - storage_path (str) - path to directory where generated files and directories will be stored.
   """
 
@@ -239,10 +240,18 @@ class BlueprintGenerator:
     else:
       a3_megagpu_pool_0.update_settings({"static_node_count": num_nodes})
 
+    if capacity_type not in (CapacityType.SPOT, CapacityType.FLEX_START):
+      a3_megagpu_pool_0.update_settings(
+          {"placement_policy": {"type": "COMPACT"}}
+      )
+
     if release_channel == ReleaseChannel.RAPID:
       a3_megagpu_pool_0.set_setting("auto_upgrade", True)
 
-    set_placement_policy = capacity_type != CapacityType.SPOT
+    set_placement_policy = capacity_type not in (
+        CapacityType.SPOT,
+        CapacityType.FLEX_START,
+    )
     workload = DeploymentModule(
         id="workload_component_install",
         source="modules/management/kubectl-apply",
@@ -521,7 +530,7 @@ class BlueprintGenerator:
         settings={
             "release_channel": release_channel.value,
             "version_prefix": version_prefix,
-            "min_cluster_version": cluster_version,
+            "min_master_version": cluster_version,
             "prefix_with_deployment_name": False,
             "name_suffix": cluster_name,
             "system_node_pool_machine_type": system_node_pool_machine_type,
@@ -614,6 +623,7 @@ class BlueprintGenerator:
       gpu_pool.update_settings(self.get_dws_flex_start())
     else:
       gpu_pool.update_settings({"static_node_count": num_nodes})
+      gpu_pool.update_settings({"placement_policy": {"type": "COMPACT"}})
 
     if release_channel == ReleaseChannel.RAPID:
       gpu_pool.set_setting("auto_upgrade", True)
@@ -809,7 +819,7 @@ class BlueprintGenerator:
         settings={
             "release_channel": release_channel.value,
             "version_prefix": version_prefix,
-            "min_cluster_version": cluster_version,
+            "min_master_version": cluster_version,
             "system_node_pool_machine_type": system_node_pool_machine_type,
             "system_node_pool_node_count": {
                 "total_min_nodes": system_node_pool_min_node_count,
@@ -896,6 +906,7 @@ class BlueprintGenerator:
       gpu_pool.update_settings(self.get_dws_flex_start())
     else:
       gpu_pool.update_settings({"static_node_count": num_nodes})
+      gpu_pool.update_settings({"placement_policy": {"type": "COMPACT"}})
 
     if release_channel == ReleaseChannel.RAPID:
       gpu_pool.set_setting("auto_upgrade", True)

xpk/core/blueprint/testing/data/a3_ultra.yaml

@@ -97,7 +97,7 @@ deployment_groups:
     settings:
       release_channel: RAPID
       version_prefix: '1.2'
-      min_cluster_version: 1.2.3
+      min_master_version: 1.2.3
       prefix_with_deployment_name: false
       name_suffix: gke-a3-ultra
       system_node_pool_machine_type: "e2-standard-16"
@@ -142,6 +142,8 @@ deployment_groups:
         specific_reservations:
         - name: test-reservation
       static_node_count: 2
+      placement_policy:
+        type: COMPACT
     outputs: [instructions]
 
   - !DeploymentModule

xpk/core/blueprint/testing/data/a4.yaml

@@ -121,7 +121,7 @@ deployment_groups:
         network_tier=null}], ipv6_access_config=[], alias_ip_range=[]}], gke-a4-rdma-net.subnetwork_interfaces_gke))
       release_channel: RAPID
       version_prefix: '1.2'
-      min_cluster_version: 1.2.3
+      min_master_version: 1.2.3
     use:
     - gke-a4-net-0
   - !DeploymentModule
@@ -154,6 +154,8 @@ deployment_groups:
         network_ip=null, stack_type=null, access_config=[{nat_ip=null, public_ptr_domain_name=null,
         network_tier=null}], ipv6_access_config=[], alias_ip_range=[]}], gke-a4-rdma-net.subnetwork_interfaces_gke))
       static_node_count: 2
+      placement_policy:
+        type: COMPACT
 
   - !DeploymentModule
     id: workload-manager-install

xpk/core/capacity.py

@@ -90,7 +90,7 @@ def get_capacity_type(args) -> tuple[CapacityType, int]:
     capacity_type = CapacityType.ON_DEMAND
     num_types += 1
   if args.reservation:
-    return_code = verify_reservation_exists(args)
+    return_code = verify_reservations_exist(args)
     if return_code > 0:
       return capacity_type, return_code
     capacity_type = CapacityType.RESERVATION
@@ -184,8 +184,22 @@ def get_reservation_deployment_type(
   return output.strip()
 
 
-def verify_reservation_exists(args) -> int:
-  """Verify the reservation exists.
+def get_reservations_list(args) -> list[str]:
+  """Get the list of reservations from args.
+
+  Args:
+    args: user provided arguments.
+
+  Returns:
+    List of strings of reservations.
+  """
+  if not args.reservation:
+    return []
+  return [r.strip() for r in args.reservation.split(',')]
+
+
+def verify_reservations_exist(args) -> int:
+  """Verify the reservations exist.
 
   Args:
     args: user provided arguments for running the command.
@@ -193,16 +207,20 @@ def verify_reservation_exists(args) -> int:
   Returns:
     0 if successful and 1 otherwise.
   """
-  reservation = parse_reservation(args.reservation, args.project)
-  command = (
-      f'gcloud beta compute reservations describe {reservation.name}'
-      f' --project={reservation.project} --zone={args.zone}'
-  )
-  return_code = run_command_with_updates(command, 'Describe reservation')
-  if return_code != 0:
-    xpk_print(f'Describe reservation returned ERROR {return_code}')
-    xpk_print('Please confirm that your reservation name is correct.')
-    return 1
+  for reservation_name in get_reservations_list(args):
+    reservation = parse_reservation(reservation_name, args.project)
+    command = (
+        f'gcloud beta compute reservations describe {reservation.name}'
+        f' --project={reservation.project} --zone={args.zone}'
+    )
+    return_code = run_command_with_updates(command, 'Describe reservation')
+    if return_code != 0:
+      xpk_print(f'Describe reservation returned ERROR {return_code}')
+      xpk_print(
+          f'Please confirm that your reservation name {reservation_name} is'
+          ' correct.'
+      )
+      return 1
   return 0
 
 
@@ -211,6 +229,7 @@ def get_capacity_arguments_from_capacity_type(
     capacity_type: CapacityType,
     max_nodes: int,
     accelerator_type: AcceleratorType,
+    reservation_name: str | None,
 ) -> tuple[str, int]:
   """Determine the Nodepool creation capacity arguments needed.
 
@@ -240,7 +259,7 @@ def get_capacity_arguments_from_capacity_type(
         capacity_args += ' --enable-queued-provisioning'
     case CapacityType.RESERVATION:
       capacity_args = (
-          f'--reservation-affinity=specific --reservation={args.reservation}'
+          f'--reservation-affinity=specific --reservation={reservation_name}'
       )
     case _:
       xpk_print(
@@ -252,13 +271,14 @@ def get_capacity_arguments_from_capacity_type(
 
 
 def get_capacity_node_selectors_from_capacity_type(
-    args, capacity_type: str
+    capacity_type: str, reservation_name: str | None
 ) -> tuple[str, int]:
   """Determine the node selectors for a workload to run on a specific capacity type.
 
   Args:
-    args: user provided arguments for running the command.
     capacity_type: The type of capacity the user configured.
+    reservation_name: The name of the reservation to use. Set to None if not
+      using reservations.
 
   Returns:
     Tuple with string with the node selectors to use and
@@ -275,7 +295,7 @@ def get_capacity_node_selectors_from_capacity_type(
     case CapacityType.SPOT.name:
       node_selector = 'cloud.google.com/gke-spot: "true"'
     case CapacityType.RESERVATION.name:
-      node_selector = f'cloud.google.com/reservation-name: {args.reservation}'
+      node_selector = f'cloud.google.com/reservation-name: {reservation_name}'
     case _:
       xpk_print(
           f'Unknown capacity type: {capacity_type}. Unable to determine the'

xpk/core/capacity_test.py

@@ -16,7 +16,15 @@ limitations under the License.
 
 import pytest
 from unittest.mock import MagicMock, patch
-from .capacity import get_reservation_deployment_type, parse_reservation, Reservation
+from .capacity import (
+    get_reservation_deployment_type,
+    parse_reservation,
+    Reservation,
+    get_capacity_type,
+    CapacityType,
+    verify_reservations_exist,
+    get_reservations_list,
+)
 
 
 @patch('xpk.core.capacity.xpk_print')
@@ -133,3 +141,60 @@ def test_parse_reservation_fails_on_invalid_reservations(
     parse_reservation(reservation_path, 'cluster-project')
 
   assert 'Unable to parse reservation' in xpk_print.mock_calls[0].args[0]
+
+
+def test_get_capacity_type_multiple_reservations(mocker):
+  args = MagicMock()
+  args.on_demand = False
+  args.spot = False
+  args.flex = False
+  args.reservation = 'res1,res2'
+  args.project = 'test-project'
+  args.zone = 'us-central1-a'
+  mocker.patch('xpk.core.capacity.run_command_with_updates', return_value=0)
+
+  capacity_type, return_code = get_capacity_type(args)
+
+  assert capacity_type == CapacityType.RESERVATION
+  assert return_code == 0
+
+
+def test_verify_reservations_exist_multiple(mocker):
+  args = MagicMock()
+  args.reservation = 'res1,res2'
+  args.project = 'test-project'
+  args.zone = 'us-central1-a'
+
+  mock_run = mocker.patch(
+      'xpk.core.capacity.run_command_with_updates', return_value=0
+  )
+
+  return_code = verify_reservations_exist(args)
+
+  assert return_code == 0
+  assert mock_run.call_count == 2
+
+
+def test_get_reservations_list_with_single_reservation(mocker):
+  args = mocker.Mock(reservation='res1')
+  assert get_reservations_list(args) == ['res1']
+
+
+def test_get_reservations_list_with_multiple_reservations(mocker):
+  args = mocker.Mock(reservation='res1,res2')
+  assert get_reservations_list(args) == ['res1', 'res2']
+
+
+def test_get_reservations_list_with_whitespace(mocker):
+  args = mocker.Mock(reservation='res1, res2 ')
+  assert get_reservations_list(args) == ['res1', 'res2']
+
+
+def test_get_reservations_list_none(mocker):
+  args = mocker.Mock(reservation=None)
+  assert get_reservations_list(args) == []
+
+
+def test_get_reservations_list_empty(mocker):
+  args = mocker.Mock(reservation='')
+  assert get_reservations_list(args) == []

xpk/core/cluster.py

@@ -158,7 +158,7 @@ def install_nri_on_cluster() -> int:
 
 
 def get_cluster_nodes_info() -> list[dict]:
-  """Get list of cluster's nodes descrition in yaml format
+  """Get list of cluster's nodes description in yaml format
 
   Returns:
     List of nodes info yaml objects.
@@ -393,11 +393,13 @@ def project_id_to_project_number(project_id: str) -> str:
 def setup_k8s_env(args) -> k8s_client.ApiClient:
   add_zone_and_project(args)
   get_cluster_credentials(args)
-  args.project_number = (
-      project_id_to_project_number(args.project)
-      if not args.dry_run
-      else abs(hash(args.project) % (10**12))  # 12 digit hash
-  )
+  # Use provided project number if available, otherwise fetch via API
+  if getattr(args, 'project_number', None):
+    xpk_print(f'Using provided project number: {args.project_number}')
+  elif args.dry_run:
+    args.project_number = abs(hash(args.project) % (10**12))  # 12 digit hash
+  else:
+    args.project_number = project_id_to_project_number(args.project)
 
   config.load_kube_config()
   return k8s_client.ApiClient()
@@ -716,10 +718,8 @@ def get_cluster_credentials(args) -> int:
       location=location,
       dns_endpoint=True,
   )
-  if return_code != 0:
-    return return_code
 
-  if not _are_credentials_valid():
+  if return_code != 0 or not _are_credentials_valid():
     xpk_print('Detected error. Retrying without --dns-endpoint flag...')
     return_code = _get_credentials(
         project=args.project,
@@ -751,6 +751,6 @@ def _get_credentials(
 def _are_credentials_valid() -> bool:
   kubectl_command = 'kubectl get pods'
   kubectl_return_code = run_command_with_updates(
-      kubectl_command, 'Test kubectl credentials'
+      kubectl_command, 'Test kubectl credentials', verbose=False
   )
   return kubectl_return_code == 0