Fix traceback on timeout with slicing + facts (#13139)

Slicing a QS with a step parameter forces the QS and returns a list.

Fixes #13131

Signed-off-by: Rick Elrod <rick@elrod.me>

.github/workflows/pr_body_check.yml

@@ -13,21 +13,13 @@ jobs:
       packages: write
       contents: read
     steps:
-      - name: Write PR body to a file
-        run: |
-          cat >> pr.body << __SOME_RANDOM_PR_EOF__
-          ${{ github.event.pull_request.body }}
-          __SOME_RANDOM_PR_EOF__
-
-      - name: Display the received body for troubleshooting
-        run: cat pr.body
-
-      # We want to write these out individually just incase the options were joined on a single line
       - name: Check for each of the lines
+        env:
+          PR_BODY: ${{ github.event.pull_request.body }}
         run: |
-          grep "Bug, Docs Fix or other nominal change" pr.body > Z
+          echo $PR_BODY | grep "Bug, Docs Fix or other nominal change" > Z
-          grep "New or Enhanced Feature" pr.body > Y
+          echo $PR_BODY | grep "New or Enhanced Feature" > Y
-          grep "Breaking Change" pr.body > X
+          echo $PR_BODY | grep "Breaking Change" > X
           exit 0
         # We exit 0 and set the shell to prevent the returns from the greps from failing this step
         # See https://docs.github.com/en/actions/using-workflows/workflow-syntax-for-github-actions#exit-codes-and-error-action-preference

Makefile

@@ -85,6 +85,7 @@ clean: clean-ui clean-api clean-awxkit clean-dist
 
 clean-api:
 	rm -rf build $(NAME)-$(VERSION) *.egg-info
+	rm -rf .tox
 	find . -type f -regex ".*\.py[co]$$" -delete
 	find . -type d -name "__pycache__" -delete
 	rm -f awx/awx_test.sqlite3*
@@ -181,7 +182,7 @@ collectstatic:
 	@if [ "$(VENV_BASE)" ]; then \
 		. $(VENV_BASE)/awx/bin/activate; \
 	fi; \
-	mkdir -p awx/public/static && $(PYTHON) manage.py collectstatic --clear --noinput > /dev/null 2>&1
+	$(PYTHON) manage.py collectstatic --clear --noinput > /dev/null 2>&1
 
 DEV_RELOAD_COMMAND ?= supervisorctl restart tower-processes:*
 
@@ -377,6 +378,8 @@ clean-ui:
 	rm -rf awx/ui/build
 	rm -rf awx/ui/src/locales/_build
 	rm -rf $(UI_BUILD_FLAG_FILE)
+        # the collectstatic command doesn't like it if this dir doesn't exist.
+	mkdir -p awx/ui/build/static
 
 awx/ui/node_modules:
 	NODE_OPTIONS=--max-old-space-size=6144 $(NPM_BIN) --prefix awx/ui --loglevel warn --force ci
@@ -386,16 +389,14 @@ $(UI_BUILD_FLAG_FILE):
 	$(PYTHON) tools/scripts/compilemessages.py
 	$(NPM_BIN) --prefix awx/ui --loglevel warn run compile-strings
 	$(NPM_BIN) --prefix awx/ui --loglevel warn run build
-	mkdir -p awx/public/static/css
+	mkdir -p /var/lib/awx/public/static/css
-	mkdir -p awx/public/static/js
+	mkdir -p /var/lib/awx/public/static/js
-	mkdir -p awx/public/static/media
+	mkdir -p /var/lib/awx/public/static/media
-	cp -r awx/ui/build/static/css/* awx/public/static/css
+	cp -r awx/ui/build/static/css/* /var/lib/awx/public/static/css
-	cp -r awx/ui/build/static/js/* awx/public/static/js
+	cp -r awx/ui/build/static/js/* /var/lib/awx/public/static/js
-	cp -r awx/ui/build/static/media/* awx/public/static/media
+	cp -r awx/ui/build/static/media/* /var/lib/awx/public/static/media
 	touch $@
 
-
-
 ui-release: $(UI_BUILD_FLAG_FILE)
 
 ui-devel: awx/ui/node_modules
@@ -453,6 +454,7 @@ COMPOSE_OPTS ?=
 CONTROL_PLANE_NODE_COUNT ?= 1
 EXECUTION_NODE_COUNT ?= 2
 MINIKUBE_CONTAINER_GROUP ?= false
+MINIKUBE_SETUP ?= false # if false, run minikube separately
 EXTRA_SOURCES_ANSIBLE_OPTS ?=
 
 ifneq ($(ADMIN_PASSWORD),)
@@ -461,7 +463,7 @@ endif
 
 docker-compose-sources: .git/hooks/pre-commit
 	@if [ $(MINIKUBE_CONTAINER_GROUP) = true ]; then\
-	    ansible-playbook -i tools/docker-compose/inventory tools/docker-compose-minikube/deploy.yml; \
+	    ansible-playbook -i tools/docker-compose/inventory -e minikube_setup=$(MINIKUBE_SETUP) tools/docker-compose-minikube/deploy.yml; \
 	fi;
 
 	ansible-playbook -i tools/docker-compose/inventory tools/docker-compose/ansible/sources.yml \
@@ -635,4 +637,4 @@ help/generate:
 		} \
 	} \
 	{ lastLine = $$0 }' $(MAKEFILE_LIST) | sort -u
-	@printf "\n"
+	@printf "\n"

awx/api/serializers.py

@@ -4952,7 +4952,7 @@ class InstanceSerializer(BaseSerializer):
             res['install_bundle'] = self.reverse('api:instance_install_bundle', kwargs={'pk': obj.pk})
         res['peers'] = self.reverse('api:instance_peers_list', kwargs={"pk": obj.pk})
         if self.context['request'].user.is_superuser or self.context['request'].user.is_system_auditor:
-            if obj.node_type != 'hop':
+            if obj.node_type == 'execution':
                 res['health_check'] = self.reverse('api:instance_health_check', kwargs={'pk': obj.pk})
         return res
 

awx/api/views/__init__.py

@@ -392,8 +392,8 @@ class InstanceHealthCheck(GenericAPIView):
     permission_classes = (IsSystemAdminOrAuditor,)
 
     def get_queryset(self):
+        return super().get_queryset().filter(node_type='execution')
         # FIXME: For now, we don't have a good way of checking the health of a hop node.
-        return super().get_queryset().exclude(node_type='hop')
 
     def get(self, request, *args, **kwargs):
         obj = self.get_object()
@@ -413,9 +413,10 @@ class InstanceHealthCheck(GenericAPIView):
 
             execution_node_health_check.apply_async([obj.hostname])
         else:
-            from awx.main.tasks.system import cluster_node_health_check
+            return Response(
-
+                {"error": f"Cannot run a health check on instances of type {obj.node_type}.  Health checks can only be run on execution nodes."},
-            cluster_node_health_check.apply_async([obj.hostname], queue=obj.hostname)
+                status=status.HTTP_400_BAD_REQUEST,
+            )
         return Response({'msg': f"Health check is running for {obj.hostname}."}, status=status.HTTP_200_OK)
 
 

awx/main/models/inventory.py

@@ -247,6 +247,19 @@ class Inventory(CommonModelNameNotUnique, ResourceMixin, RelatedJobsMixin):
         return (number, step)
 
     def get_sliced_hosts(self, host_queryset, slice_number, slice_count):
+        """
+        Returns a slice of Hosts given a slice number and total slice count, or
+        the original queryset if slicing is not requested.
+
+        NOTE: If slicing is performed, this will return a List[Host] with the
+        resulting slice. If slicing is not performed it will return the
+        original queryset (not evaluating it or forcing it to a list). This
+        puts the burden on the caller to check the resulting type. This is
+        non-ideal because it's easy to get wrong, but I think the only way
+        around it is to force the queryset which has memory implications for
+        large inventories.
+        """
+
         if slice_count > 1 and slice_number > 0:
             offset = slice_number - 1
             host_queryset = host_queryset[offset::slice_count]

awx/main/models/jobs.py

@@ -15,6 +15,7 @@ from urllib.parse import urljoin
 from django.conf import settings
 from django.core.exceptions import ValidationError
 from django.db import models
+from django.db.models.query import QuerySet
 
 # from django.core.cache import cache
 from django.utils.encoding import smart_str
@@ -844,22 +845,30 @@ class Job(UnifiedJob, JobOptions, SurveyJobMixin, JobNotificationMixin, TaskMana
     def get_notification_friendly_name(self):
         return "Job"
 
-    def _get_inventory_hosts(self, only=['name', 'ansible_facts', 'ansible_facts_modified', 'modified', 'inventory_id']):
+    def _get_inventory_hosts(self, only=('name', 'ansible_facts', 'ansible_facts_modified', 'modified', 'inventory_id'), **filters):
+        """Return value is an iterable for the relevant hosts for this job"""
         if not self.inventory:
             return []
         host_queryset = self.inventory.hosts.only(*only)
-        return self.inventory.get_sliced_hosts(host_queryset, self.job_slice_number, self.job_slice_count)
+        if filters:
+            host_queryset = host_queryset.filter(**filters)
+        host_queryset = self.inventory.get_sliced_hosts(host_queryset, self.job_slice_number, self.job_slice_count)
+        if isinstance(host_queryset, QuerySet):
+            return host_queryset.iterator()
+        return host_queryset
 
     def start_job_fact_cache(self, destination, modification_times, timeout=None):
         self.log_lifecycle("start_job_fact_cache")
         os.makedirs(destination, mode=0o700)
-        hosts = self._get_inventory_hosts()
+
         if timeout is None:
             timeout = settings.ANSIBLE_FACT_CACHE_TIMEOUT
         if timeout > 0:
             # exclude hosts with fact data older than `settings.ANSIBLE_FACT_CACHE_TIMEOUT seconds`
             timeout = now() - datetime.timedelta(seconds=timeout)
-            hosts = hosts.filter(ansible_facts_modified__gte=timeout)
+            hosts = self._get_inventory_hosts(ansible_facts_modified__gte=timeout)
+        else:
+            hosts = self._get_inventory_hosts()
         for host in hosts:
             filepath = os.sep.join(map(str, [destination, host.name]))
             if not os.path.realpath(filepath).startswith(destination):

awx/main/tasks/receptor.py

@@ -208,7 +208,10 @@ def run_until_complete(node, timing_data=None, **kwargs):
     if state_name.lower() == 'failed':
         work_detail = status.get('Detail', '')
         if work_detail:
-            raise RemoteJobError(f'Receptor error from {node}, detail:\n{work_detail}')
+            if stdout:
+                raise RemoteJobError(f'Receptor error from {node}, detail:\n{work_detail}\nstdout:\n{stdout}')
+            else:
+                raise RemoteJobError(f'Receptor error from {node}, detail:\n{work_detail}')
         else:
             raise RemoteJobError(f'Unknown ansible-runner error on node {node}, stdout:\n{stdout}')
 

awx/main/tests/functional/api/test_instance.py

@@ -7,7 +7,7 @@ from awx.main.models.ha import Instance
 from django.test.utils import override_settings
 
 
-INSTANCE_KWARGS = dict(hostname='example-host', cpu=6, memory=36000000000, cpu_capacity=6, mem_capacity=42)
+INSTANCE_KWARGS = dict(hostname='example-host', cpu=6, node_type='execution', memory=36000000000, cpu_capacity=6, mem_capacity=42)
 
 
 @pytest.mark.django_db

awx/main/utils/handlers.py

@@ -110,7 +110,7 @@ if settings.COLOR_LOGS is True:
                 # logs rendered with cyan text
                 previous_level_map = self.level_map.copy()
                 if record.name == "awx.analytics.job_lifecycle":
-                    self.level_map[logging.DEBUG] = (None, 'cyan', True)
+                    self.level_map[logging.INFO] = (None, 'cyan', True)
                 msg = super(ColorHandler, self).colorize(line, record)
                 self.level_map = previous_level_map
                 return msg

awx/settings/defaults.py

@@ -101,7 +101,7 @@ USE_L10N = True
 
 USE_TZ = True
 
-STATICFILES_DIRS = (os.path.join(BASE_DIR, 'ui', 'build', 'static'), os.path.join(BASE_DIR, 'static'))
+STATICFILES_DIRS = [os.path.join(BASE_DIR, 'ui', 'build', 'static'), os.path.join(BASE_DIR, 'static')]
 
 # Absolute filesystem path to the directory where static file are collected via
 # the collectstatic command.

awx/ui/src/components/Schedule/shared/ScheduleForm.js

@@ -416,8 +416,14 @@ function ScheduleForm({
 
       if (options.end === 'onDate') {
         if (
-          DateTime.fromISO(values.startDate) >=
+          DateTime.fromFormat(
-          DateTime.fromISO(options.endDate)
+            `${values.startDate} ${values.startTime}`,
+            'yyyy-LL-dd h:mm a'
+          ).toMillis() >=
+          DateTime.fromFormat(
+            `${options.endDate} ${options.endTime}`,
+            'yyyy-LL-dd h:mm a'
+          ).toMillis()
         ) {
           freqErrors.endDate = t`Please select an end date/time that comes after the start date/time.`;
         }

awx/ui/src/components/Schedule/shared/ScheduleForm.test.js

@@ -900,6 +900,36 @@ describe('<ScheduleForm />', () => {
       );
     });
 
+    test('should create schedule with the same start and end date provided that the end date is at a later time', async () => {
+      const today = DateTime.now().toFormat('yyyy-LL-dd');
+      const laterTime = DateTime.now().plus({ hours: 1 }).toFormat('h:mm a');
+      await act(async () => {
+        wrapper.find('DatePicker[aria-label="End date"]').prop('onChange')(
+          today,
+          new Date(today)
+        );
+      });
+      wrapper.update();
+      expect(
+        wrapper
+          .find('FormGroup[data-cy="schedule-End date/time"]')
+          .prop('helperTextInvalid')
+      ).toBe(
+        'Please select an end date/time that comes after the start date/time.'
+      );
+      await act(async () => {
+        wrapper.find('TimePicker[aria-label="End time"]').prop('onChange')(
+          laterTime
+        );
+      });
+      wrapper.update();
+      expect(
+        wrapper
+          .find('FormGroup[data-cy="schedule-End date/time"]')
+          .prop('helperTextInvalid')
+      ).toBe(undefined);
+    });
+
     test('error shown when on day number is not between 1 and 31', async () => {
       await act(async () => {
         wrapper.find('FrequencySelect#schedule-frequency').invoke('onChange')([

awx_collection/plugins/modules/project_update.py

@@ -114,7 +114,12 @@ def main():
     # Update the project
     result = module.post_endpoint(project['related']['update'])
 
-    if result['status_code'] != 202:
+    if result['status_code'] == 405:
+        module.fail_json(
+            msg="Unable to trigger a project update because the project scm_type ({0}) does not support it.".format(project['scm_type']),
+            response=result
+        )
+    elif result['status_code'] != 202:
         module.fail_json(msg="Failed to update project, see response for details", response=result)
 
     module.json_output['changed'] = True

awxkit/awxkit/api/pages/api.py

@@ -275,7 +275,13 @@ class ApiV2(base.Base):
                         # When creating a project, we need to wait for its
                         # first project update to finish so that associated
                         # JTs have valid options for playbook names
-                        _page.wait_until_completed()
+                        try:
+                            _page.wait_until_completed(timeout=300)
+                        except AssertionError:
+                            # If the project update times out, try to
+                            # carry on in the hopes that it will
+                            # finish before it is needed.
+                            pass
                 else:
                     # If we are an existing project and our scm_tpye is not changing don't try and import the local_path setting
                     if asset['natural_key']['type'] == 'project' and 'local_path' in post_data and _page['scm_type'] == post_data['scm_type']:

awxkit/awxkit/api/pages/schedules.py

@@ -1,6 +1,7 @@
 from contextlib import suppress
 
-from awxkit.api.pages import SystemJobTemplate
+from awxkit.api.pages import JobTemplate, SystemJobTemplate, Project, InventorySource
+from awxkit.api.pages.workflow_job_templates import WorkflowJobTemplate
 from awxkit.api.mixins import HasCreate
 from awxkit.api.resources import resources
 from awxkit.config import config
@@ -11,7 +12,7 @@ from . import base
 
 
 class Schedule(HasCreate, base.Base):
-    dependencies = [SystemJobTemplate]
+    dependencies = [JobTemplate, SystemJobTemplate, Project, InventorySource, WorkflowJobTemplate]
     NATURAL_KEY = ('unified_job_template', 'name')
 
     def silent_delete(self):

tools/ansible/roles/dockerfile/defaults/main.yml

@@ -9,4 +9,4 @@ template_dest: '_build'
 receptor_image: quay.io/ansible/receptor:devel
 
 # Helper vars to construct the proper download URL for the current architecture
-image_architecture: '{{ { "x86_64": "amd64", "aarch64": "arm64", "armv7": "arm", "ppc64le": "ppc64le" }[ansible_facts.architecture] }}'
+image_architecture: '{{ { "x86_64": "amd64", "aarch64": "arm64", "armv7": "arm", "arm64": "arm64", "ppc64le": "ppc64le" }[ansible_facts.architecture] }}'

tools/docker-compose-minikube/minikube/defaults/main.yml

@@ -9,8 +9,8 @@ addons:
 minikube_url_linux: 'https://storage.googleapis.com/minikube/releases/latest/minikube-linux-amd64'
 minikube_url_macos: 'https://storage.googleapis.com/minikube/releases/latest/minikube-darwin-amd64'
 
-kubectl_url_linux: 'https://dl.k8s.io/release/v1.21.0/bin/linux/amd64/kubectl'
+kubectl_url_linux: 'https://dl.k8s.io/release/v1.25.0/bin/linux/amd64/kubectl'
-kubectl_url_macos: 'https://dl.k8s.io/release/v1.21.0/bin/darwin/amd64/kubectl'
+kubectl_url_macos: 'https://dl.k8s.io/release/v1.25.0/bin/darwin/amd64/kubectl'
 
 # Service Account Name
 minikube_service_account_name: 'awx-devel'

tools/docker-compose-minikube/minikube/tasks/main.yml

@@ -8,6 +8,10 @@
     state: 'directory'
     mode: '0700'
 
+- name: debug minikube_setup
+  debug:
+    var: minikube_setup
+
 # Linux block
 - block:
     - name: Download Minikube
@@ -24,6 +28,7 @@
   when:
     - ansible_architecture == "x86_64"
     - ansible_system == "Linux"
+    - minikube_setup | default(False) | bool
 
 # MacOS block
 - block:
@@ -41,25 +46,29 @@
   when:
     - ansible_architecture == "x86_64"
     - ansible_system == "Darwin"
+    - minikube_setup | default(False) | bool
 
-- name: Starting Minikube
+- block:
-  shell: "{{ sources_dest }}/minikube start --driver={{ driver }} --install-addons=true --addons={{ addons | join(',') }}"
+    - name: Starting Minikube
-  register: minikube_stdout
+      shell: "{{ sources_dest }}/minikube start --driver={{ driver }} --install-addons=true --addons={{ addons | join(',') }}"
+      register: minikube_stdout
 
-- name: Enable Ingress Controller on Minikube
+    - name: Enable Ingress Controller on Minikube
-  shell: "{{ sources_dest }}/minikube addons enable ingress"
+      shell: "{{ sources_dest }}/minikube addons enable ingress"
+      when:
+        - minikube_stdout.rc == 0
+      register: _minikube_ingress
+      ignore_errors: true
+
+    - name: Show Minikube Ingress known-issue 7332 warning
+      pause:
+        seconds: 5
+        prompt: "The Minikube Ingress addon has been disabled since it looks like you are hitting https://github.com/kubernetes/minikube/issues/7332"
+      when:
+        - '"minikube/issues/7332" in _minikube_ingress.stderr'
+        - ansible_system == "Darwin"
   when:
-    - minikube_stdout.rc == 0
+    - minikube_setup | default(False) | bool
-  register: _minikube_ingress
-  ignore_errors: true
-
-- name: Show Minikube Ingress known-issue 7332 warning
-  pause:
-    seconds: 5
-    prompt: "The Minikube Ingress addon has been disabled since it looks like you are hitting https://github.com/kubernetes/minikube/issues/7332"
-  when:
-    - '"minikube/issues/7332" in _minikube_ingress.stderr'
-    - ansible_system == "Darwin"
 
 - name: Create ServiceAccount and clusterRoleBinding
   k8s:

tools/docker-compose/README.md

@@ -301,11 +301,19 @@ Note that you may see multiple messages of the form `2021-03-04 20:11:47,666 WAR
 
 To bring up a 1 node AWX + minikube that is accessible from AWX run the following.
 
+Start minikube
+
+```bash
+(host)$minikube start --cpus=4  --memory=8g --addons=ingress`
+```
+
+Start AWX
+
 ```bash
 (host)$ make docker-compose-container-group
 ```
 
-Alternatively, you can set the env var `MINIKUBE_CONTAINER_GROUP=true` to use the default dev env bring up. his way you can use other env flags like the cluster node count.
+Alternatively, you can set the env var `MINIKUBE_CONTAINER_GROUP=true` to use the default dev env bring up. his way you can use other env flags like the cluster node count. Set `MINIKUBE_SETUP=true` to make the roles download, install and run minikube for you, but if you run into issues with this just start minikube yourself.
 
 ```bash
 (host)$ MINIKUBE_CONTAINER_GROUP=true make docker-compose

tools/docker-compose/bootstrap_development.sh

@@ -19,6 +19,9 @@ else
     wait-for-migrations
 fi
 
+# Make sure that the UI static file directory exists, Django complains otherwise.
+mkdir -p /awx_devel/awx/ui/build/static
+
 if output=$(awx-manage createsuperuser --noinput --username=admin --email=admin@localhost 2> /dev/null); then
     echo $output
 fi
@@ -27,10 +30,6 @@ echo "Admin password: ${DJANGO_SUPERUSER_PASSWORD}"
 awx-manage create_preload_data
 awx-manage register_default_execution_environments
 
-mkdir -p /awx_devel/awx/public/static
-mkdir -p /awx_devel/awx/ui/static
-mkdir -p /awx_devel/awx/ui/build/static
-
 awx-manage provision_instance --hostname="$(hostname)" --node_type="$MAIN_NODE_TYPE"
 awx-manage register_queue --queuename=controlplane --instance_percent=100
 awx-manage register_queue --queuename=default --instance_percent=100