asg_management_fix_and_improve

roles/asg_management/defaults/main.yml

@@ -2,6 +2,8 @@
 # AWS ASG variables to allow for the suspension of autoscaling during a code deployment.
 asg_management:
   name: "" # if the deploy is on an ASG put the name here
+  #target_group_name: "example" # matches the ASG name by default, specify if your TargetGroup name is different (for example due to the 32-char name length limit in AWS)
+  refresh_asg_instances: true # runs only if squashFS image unmount failed and this set to true.
   #profile: "example" # optional, the boto profile name to use if not the system default
   region: "eu-west-1"
   suspend_processes: "Launch Terminate HealthCheck" # space separated string, see https://docs.aws.amazon.com/autoscaling/ec2/userguide/as-suspend-resume-processes.html

roles/asg_management/tasks/main.yml

@@ -11,7 +11,7 @@
         region: "{{ asg_management.region }}"
         profile: "{{ asg_management.profile | default(omit) }}"
         names:
-          - "{{ asg_management.name }}"
+          - "{{ asg_management.target_group_name | default(asg_management.name) }}"
       register: _target_group
 
     - name: Loop over target instances until they are all 'healthy'.

roles/deploy_code/tasks/cleanup.yml

@@ -150,93 +150,147 @@
       failed_when: false
       register: _deploy_code_mount_check
 
-    - name: Get the current pts session.
-      ansible.builtin.shell:
-        cmd: "tty | sed 's#/dev/##'"
-      register: deploy_pts
-
-    - name: "Check for active sessions in {{ deploy_base_path }}."
-      ansible.builtin.shell:
-        cmd: "ps -eo pid,tty | awk '{print $1}' | xargs -n 1 pwdx 2>&1 | grep -v 'No such process' | grep {{ deploy_base_path }} | cut -d: -f1 | xargs -n 1 ps -o tty= -p | sort | uniq"
-      register: sessions_in_deploy_path
-      become: true
-
-    - name: Display active sessions.
-      ansible.builtin.debug:
-        msg: >
-          Deploy session: {{ deploy_pts.stdout | default('Unknown') }}.
-          Active sessions in {{ deploy_base_path }}: {{ sessions_in_deploy_path.stdout_lines | default([]) | join(', ') | default('None') }}.
-
-    - name: Kill sessions except the current one.
+    - name: Mount SquashFS image if there is no mounted one.
       ansible.builtin.command:
-        cmd: "pkill -9 -t {{ item }}"
-      loop: "{{ sessions_in_deploy_path.stdout_lines }}"
-      when:
-        - "item != deploy_pts.stdout"
-        - "item is match('^pts/\\d+$')"
-      failed_when: false
-      register: kill_sessions_result
-      become: true
-
-    - name: Display killed sessions.
-      ansible.builtin.debug:
-        msg: >
-          Sessions terminated: {{ kill_sessions_result.results | selectattr('rc', 'defined') | selectattr('rc', 'equalto', 0) | map(attribute='item') | list | join(', ') | default('None') }}.
-
-    - name: Reload any services that might be keeping the loop device busy.
-      ansible.builtin.service:
-        name: "{{ www_service }}"
-        state: reloaded
-      with_items: "{{ deploy_code.services }}"
-      loop_control:
-        loop_var: www_service
-      become: true
-      when:
-        - _deploy_code_mount_check.rc == 0
-        - deploy_code.service_action == "reload"
-        - deploy_code.services | length > 0
-
-    - name: Stop any services that might be keeping the loop device busy.
-      ansible.builtin.service:
-        name: "{{ www_service }}"
-        state: stopped
-      with_items: "{{ deploy_code.services }}"
-      loop_control:
-        loop_var: www_service
+        cmd: "mount {{ build_base_path }}/deploy.sqsh {{ deploy_base_path }} -t squashfs -o loop"
       become: true
       when:
-        - _deploy_code_mount_check.rc == 0
-        - deploy_code.service_action == "stop"
-        - deploy_code.services | length > 0
+        - _deploy_code_mount_check.rc != 0
 
-    - name: Unmount existing SquashFS image.
-      ansible.builtin.command:
-        cmd: "umount {{ deploy_base_path }}"
-      become: true
+    - name: Mount new SquashFS image instead of the already mounted.
       when:
         - _deploy_code_mount_check.rc == 0
-      register: task_result
-      retries: "{{ deploy_code.unmount_retries }}"
-      delay: "{{ deploy_code.unmount_delay }}"
-      until: task_result.rc == 0
+      block:
+        - name: Get the current pts session.
+          ansible.builtin.shell:
+            cmd: "tty | sed 's#/dev/##'"
+          register: deploy_pts
 
-    - name: Mount new SquashFS image.
-      ansible.builtin.command:
-        cmd: "mount {{ build_base_path }}/deploy.sqsh {{ deploy_base_path }} -t squashfs -o loop"
-      become: true
+        - name: "Check for active sessions in {{ deploy_base_path }}."
+          ansible.builtin.shell:
+            cmd: "ps -eo pid,tty | awk '{print $1}' | xargs -n 1 pwdx 2>&1 | grep -v 'No such process' | grep {{ deploy_base_path }} | cut -d: -f1 | xargs -n 1 ps -o tty= -p | sort | uniq"
+          register: sessions_in_deploy_path
+          become: true
 
-    - name: Start any services we stopped.
-      ansible.builtin.service:
-        name: "{{ www_service }}"
-        state: started
-      with_items: "{{ deploy_code.services }}"
-      loop_control:
-        loop_var: www_service
-      become: true
-      when:
-        - _deploy_code_mount_check.rc == 0
-        - deploy_code.service_action == "stop"
-        - deploy_code.services | length > 0
+        - name: Display active sessions.
+          ansible.builtin.debug:
+            msg: >
+              Deploy session: {{ deploy_pts.stdout | default('Unknown') }}.
+              Active sessions in {{ deploy_base_path }}: {{ sessions_in_deploy_path.stdout_lines | default([]) | join(', ') | default('None') }}.
+
+        - name: Kill sessions except the current one.
+          ansible.builtin.command:
+            cmd: "pkill -9 -t {{ item }}"
+          loop: "{{ sessions_in_deploy_path.stdout_lines }}"
+          when:
+            - "item != deploy_pts.stdout"
+            - "item is match('^pts/\\d+$')"
+          failed_when: false
+          register: kill_sessions_result
+          become: true
+
+        - name: Display killed sessions.
+          ansible.builtin.debug:
+            msg: >
+              Sessions terminated: {{ kill_sessions_result.results | selectattr('rc', 'defined') | selectattr('rc', 'equalto', 0) | map(attribute='item') | list | join(', ') | default('None') }}.
+
+        - name: Reload any services that might be keeping the loop device busy.
+          ansible.builtin.service:
+            name: "{{ www_service }}"
+            state: reloaded
+          with_items: "{{ deploy_code.services }}"
+          loop_control:
+            loop_var: www_service
+          become: true
+          when:
+            - deploy_code.service_action == "reload"
+            - deploy_code.services | length > 0
+
+        - name: Stop any services that might be keeping the loop device busy.
+          ansible.builtin.service:
+            name: "{{ www_service }}"
+            state: stopped
+          with_items: "{{ deploy_code.services }}"
+          loop_control:
+            loop_var: www_service
+          become: true
+          when:
+            - deploy_code.service_action == "stop"
+            - deploy_code.services | length > 0
+
+        - name: Unmount existing SquashFS image.
+          ansible.builtin.command:
+            cmd: "umount {{ deploy_base_path }}"
+          become: true
+          register: task_result
+          retries: "{{ deploy_code.unmount_retries }}"
+          delay: "{{ deploy_code.unmount_delay }}"
+          until: task_result.rc == 0
+          failed_when: false
+
+        - name: If current image unmount succeeded.
+          when: task_result.rc == 0
+          block:
+            - name: Mount new SquashFS image.
+              ansible.builtin.command:
+                cmd: "mount {{ build_base_path }}/deploy.sqsh {{ deploy_base_path }} -t squashfs -o loop"
+              become: true
+
+            - name: Start any services that we stopped if the image re-mounting was successful.
+              ansible.builtin.service:
+                name: "{{ www_service }}"
+                state: started
+              with_items: "{{ deploy_code.services }}"
+              loop_control:
+                loop_var: www_service
+              become: true
+              when:
+                - deploy_code.service_action == "stop"
+                - deploy_code.services | length > 0
+
+        - name: If current image unmount failed.
+          when: task_result.rc != 0
+          block:
+            - name: Resume all autoscale processes on ASG.
+              amazon.aws.autoscaling_group:
+                name: "{{ asg_management.name }}"
+                region: "{{ asg_management.region }}"
+                profile: "{{ asg_management.profile | default(omit) }}"
+                suspended_processes: []
+              delegate_to: localhost
+              run_once: true
+              when:
+                - asg_management.name | length > 0
+
+            - name: Run ASG instance refresh.
+              amazon.aws.autoscaling_instance_refresh:
+                name: "{{ asg_management.name }}"
+                region: "{{ asg_management.region }}"
+                profile: "{{ asg_management.profile | default(omit) }}"
+                strategy: Rolling
+                preferences:
+                  min_healthy_percentage: 51
+                  instance_warmup: 100
+                  skip_matching: false
+                state: started
+              delegate_to: localhost
+              run_once: true
+              when:
+                - asg_management.name | length > 0
+                - asg_management.refresh_asg_instances
+
+            - name: Start any services we stopped if the image re-mounting failed and ASG management is disabled.
+              ansible.builtin.service:
+                name: "{{ www_service }}"
+                state: started
+              with_items: "{{ deploy_code.services }}"
+              loop_control:
+                loop_var: www_service
+              become: true
+              when:
+                - deploy_code.service_action == "stop"
+                - deploy_code.services | length > 0
+                - not asg_management.refresh_asg_instances or (asg_management.name | length) == 0
 # End of the squashFS block.
 
 - name: Trigger an infrastructure rebuild.