fix: add missing pmie webhook action configuration functionality

README.md

@@ -49,6 +49,12 @@ Example:
 metrics_monitored_hosts: ["webserver.example.com", "database.example.com"]
 ```
 
+### metrics_webhook_endpoint: ''
+
+Webhook endpoint (URL) where notification about any automatically detected
+performance issues are to be sent.  By default, these events are logged to
+the local system log only.
+
 ### metrics_retention_days: 14
 
 Retain historical performance data for the specified number of days; after

defaults/main.yml

@@ -46,3 +46,8 @@ metrics_manage_firewall: false
 # If true, manage the pmcd port, pmproxy ports using the selinux role
 # depending upon the configuration parameters.
 metrics_manage_selinux: false
+
+# Webhook endpoint (URL) where notification about any automatically detected
+# performance issues are to be sent.  By default, these events are logged to the
+# local system log only.
+metrics_webhook_endpoint: ''

tasks/main.yml

@@ -84,6 +84,7 @@
 
 - name: Setup metric collection service.
   vars:
+    pcp_pmie_endpoint: "{{ metrics_webhook_endpoint }}"
     pcp_pmlogger_discard: "{{ metrics_retention_days }}"
     pcp_target_hosts: "{{ metrics_monitored_hosts }}"
     pcp_optional_agents: "{{ __metrics_domains }}"

tests/check_notification.yml

@@ -0,0 +1,7 @@
+# SPDX-License-Identifier: MIT
+---
+# yamllint disable rule:line-length
+- name: Check if webhook notification is set
+  command: grep -E 'webhook_endpoint .* "?{{ __test_webhook }}"?' /var/lib/pcp/config/pmie/config.default
+  changed_when: false
+# yamllint enable rule:line-length

tests/handle_test_failure.yml

@@ -0,0 +1,19 @@
+---
+- name: Collect logs
+  shell: |
+    journalctl -ex
+    echo '##################'
+    echo List of SELinux AVCs - note list may be empty
+    grep type=AVC /var/log/audit/audit.log
+    echo '##################'
+    ls -alrtF /run
+    if [ -d /run/pcp ]; then
+      ls -alrtF /run/pcp
+    else
+      echo ERROR - /run/pcp does not exist
+    fi
+  changed_when: false
+
+- name: Reraise error
+  fail:
+    msg: "{{ ansible_failed_result }}"

tests/restore_services_state.yml

@@ -16,6 +16,7 @@
   when:
     - item + '.service' in final_state.ansible_facts.services
     - item + '.service' in initial_state.ansible_facts.services
+    - initial_state.ansible_facts.services[item + '.service']['status'] != 'not-found'
   with_items:
     - pmcd
     - pmlogger

tests/tests_bz1855539.yml

@@ -5,10 +5,7 @@
   hosts: all
   vars:
     pcp_pmie_control_path: "/etc/pcp/pmie/control.d/"
-  roles:
-    - role: linux-system-roles.metrics
-      vars:
-        metrics_monitored_hosts: ["127.0.0.2", "127.0.0.3"]
+    metrics_monitored_hosts: ["127.0.0.2", "127.0.0.3"]
 
   pre_tasks:
     - name: Stop test
@@ -21,22 +18,39 @@
       import_tasks: get_services_state.yml
 
   tasks:
-    - name: Check if pmie configuration file on remote host is the secondary one
-      command: |-
-        grep -E '^\s*\S+\s+n\s+n\s+' {{ pcp_pmie_control_path }}/{{ item }}
-      loop: "{{ metrics_monitored_hosts }}"
-      changed_when: false
-
-    - name: Check if pmie configuration file on remote host has the header
-      vars:
-        __test_config_path: "{{ pcp_pmie_control_path }}{{ item }}"
-      include_tasks: check_header.yml
-      loop: "{{ metrics_monitored_hosts }}"
-
-    - name: Check if pmie configuration file on local host is the primary one
-      command: grep -E '^\s*\S+\s+y\s+n\s+'  {{ pcp_pmie_control_path }}/local
-      changed_when: false
-
-  post_tasks:
-    - name: Restore state of services
-      import_tasks: restore_services_state.yml
+    - name: Run test
+      block:
+        - name: Include role
+          include_role:
+            name: linux-system-roles.metrics
+            public: true
+
+        - name: >-
+            Check if pmie configuration file on remote host is the secondary one
+          command: |-
+            grep -E '^\s*\S+\s+n\s+n\s+' {{ pcp_pmie_control_path }}/{{ item }}
+          loop: "{{ metrics_monitored_hosts }}"
+          changed_when: false
+
+        - name: Check if pmie configuration file on remote host has the header
+          vars:
+            __test_config_path: "{{ pcp_pmie_control_path }}{{ item }}"
+          include_tasks: check_header.yml
+          loop: "{{ metrics_monitored_hosts }}"
+
+        - name: >-
+            Check if pmie configuration file on local host is the primary one
+          command: >-
+            grep -E '^\s*\S+\s+y\s+n\s+' {{ pcp_pmie_control_path }}/local
+          changed_when: false
+
+        - name: Flush handlers
+          meta: flush_handlers
+
+      rescue:
+        - name: Handle failure case
+          include_tasks: handle_test_failure.yml
+
+      always:
+        - name: Restore state of services
+          import_tasks: restore_services_state.yml

tests/tests_bz1855544.yml

@@ -3,13 +3,10 @@
 - name: "Bug 1855544 - metrics role should automate the setup of
          Grafana datasources"
   hosts: all
-
-  roles:
-    - role: linux-system-roles.metrics
-      vars:
-        metrics_query_service: true
-        metrics_graph_service: true
-        metrics_from_bpftrace: true
+  vars:
+    metrics_query_service: true
+    metrics_graph_service: true
+    metrics_from_bpftrace: true
 
   pre_tasks:
     - name: Stop test
@@ -22,17 +19,31 @@
       import_tasks: get_services_state.yml
 
   tasks:
-    - name: Check if all default datasources are configured
-      include_tasks: check_default_datasources.yml
-
-    - name: Check if all required PMDAs are installed
-      command: pmprobe -I pmcd.agent.status
-      register: status
-      until: status.stdout.find("bpftrace") != -1
-      retries: 10
-      delay: 1
-      changed_when: false
-
-  post_tasks:
-    - name: Restore state of services
-      import_tasks: restore_services_state.yml
+    - name: Run test
+      block:
+        - name: Run the role
+          include_role:
+            name: linux-system-roles.metrics
+            public: true
+
+        - name: Check if all default datasources are configured
+          include_tasks: check_default_datasources.yml
+
+        - name: Check if all required PMDAs are installed
+          command: pmprobe -I pmcd.agent.status
+          register: status
+          until: status.stdout.find("bpftrace") != -1
+          retries: 10
+          delay: 1
+          changed_when: false
+
+        - name: Flush handlers
+          meta: flush_handlers
+
+      rescue:
+        - name: Handle test failure
+          include_tasks: handle_test_failure.yml
+
+      always:
+        - name: Restore state of services
+          import_tasks: restore_services_state.yml

tests/tests_default.yml

@@ -3,5 +3,25 @@
 - name: Test the role with default params
   hosts: all
 
-  roles:
-    - linux-system-roles.metrics
+  pre_tasks:
+    - name: Save state of services
+      import_tasks: get_services_state.yml
+
+  tasks:
+    - name: Run test
+      block:
+        - name: Include role
+          include_role:
+            name: linux-system-roles.metrics
+            public: true
+
+        - name: Flush handlers
+          meta: flush_handlers
+
+      rescue:
+        - name: Handle failure case
+          include_tasks: handle_test_failure.yml
+
+      always:
+        - name: Restore state of services
+          import_tasks: restore_services_state.yml

tests/tests_verify_auth.yml

@@ -4,15 +4,11 @@
   hosts: all
   vars:
     __test_uname: pcptest
-
-  roles:
-    - role: linux-system-roles.metrics
-      vars:
-        metrics_username: "{{ __test_uname }}"
-        metrics_password: "{{ pcptest_pw }}"
-        # test to install firewall but not selinux
-        metrics_manage_firewall: true
-        metrics_manage_selinux: false
+    metrics_username: "{{ __test_uname }}"
+    metrics_password: "{{ pcptest_pw }}"
+    # test to install firewall but not selinux
+    metrics_manage_firewall: true
+    metrics_manage_selinux: false
 
   pre_tasks:
     - name: Stop test
@@ -25,17 +21,31 @@
       import_tasks: get_services_state.yml
 
   tasks:
-    - name: Restart PMCD
-      # noqa command-instead-of-module
-      shell: systemctl restart pmcd && sleep 5
-      changed_when: false
-
-    - name: Check if SASL works
-      include_tasks: "{{ item }}"
-      loop:
-        - check_sasl.yml
-        - check_firewall_selinux.yml
-
-  post_tasks:
-    - name: Restore state of services
-      import_tasks: restore_services_state.yml
+    - name: Run test
+      block:
+        - name: Run role
+          include_role:
+            name: linux-system-roles.metrics
+            public: true
+
+        - name: Restart PMCD
+          # noqa command-instead-of-module
+          shell: systemctl restart pmcd && sleep 5
+          changed_when: false
+
+        - name: Check if SASL works
+          include_tasks: "{{ item }}"
+          loop:
+            - check_sasl.yml
+            - check_firewall_selinux.yml
+
+        - name: Flush handlers
+          meta: flush_handlers
+
+      rescue:
+        - name: Handle failure case
+          include_tasks: handle_test_failure.yml
+
+      always:
+        - name: Restore state of services
+          import_tasks: restore_services_state.yml

tests/tests_verify_basic.yml

@@ -2,27 +2,38 @@
 ---
 - name: Test basic services of the role are installed and running
   hosts: all
-
-  roles:
-    - role: linux-system-roles.metrics
-      vars:
-        # test to install selinux but not firewall
-        metrics_manage_firewall: false
-        metrics_manage_selinux: true
+  vars:
+    # test to install selinux but not firewall
+    metrics_manage_firewall: false
+    metrics_manage_selinux: true
 
   pre_tasks:
     - name: Save state of services
       import_tasks: get_services_state.yml
 
   tasks:
-    - name: Check if basic metrics role setup works
-      include_tasks: "{{ item }}"
-      loop:
-        - check_pcp.yml
-        - check_pmlogger.yml
-        - check_pmie.yml
-        - check_firewall_selinux.yml
+    - name: Run test
+      block:
+        - name: Run the role
+          include_role:
+            name: linux-system-roles.metrics
+            public: true
+
+        - name: Check if basic metrics role setup works
+          include_tasks: "{{ item }}"
+          loop:
+            - check_pcp.yml
+            - check_pmlogger.yml
+            - check_pmie.yml
+            - check_firewall_selinux.yml
+
+        - name: Flush handlers
+          meta: flush_handlers
+
+      rescue:
+        - name: Handle failure case
+          include_tasks: handle_test_failure.yml
 
-  post_tasks:
-    - name: Restore state of services
-      import_tasks: restore_services_state.yml
+      always:
+        - name: Restore state of services
+          import_tasks: restore_services_state.yml