Fix pageserver configuration generation bug (#2584)

* We had an issue with `lineinfile` usage for pageserver configuration
file: if the S3 bucket related values were changed, it would have
resulted in duplicate keys, resulting in invalid toml.

So to fix the issue, we should keep the configuration in structured
format (yaml in this case) so we can always generate syntactically
correct toml.

Inventories are converted to yaml just so that it's easier to maintain
the configuration there. Another alternative would have been a separate
variable files.

* Keep the ansible collections dir, but locally installed collections
should not be tracked.

.github/ansible/.gitignore

@@ -2,3 +2,6 @@ zenith_install.tar.gz
 .zenith_current_version
 neon_install.tar.gz
 .neon_current_version
+
+collections/*
+!collections/.keep

.github/ansible/ansible.cfg

@@ -3,6 +3,7 @@
 localhost_warning = False
 host_key_checking = False
 timeout = 30
+collections_paths = ./collections
 
 [ssh_connection]
 ssh_args   = -F ./ansible.ssh.cfg

.github/ansible/deploy.yaml

@@ -14,7 +14,8 @@
       - safekeeper
 
     - name: inform about versions
-      debug: msg="Version to deploy - {{ current_version }}"
+      debug:
+        msg: "Version to deploy - {{ current_version }}"
       tags:
       - pageserver
       - safekeeper
@@ -63,15 +64,29 @@
       tags:
       - pageserver
 
-    - name: update remote storage (s3) config
-      lineinfile:
-        path: /storage/pageserver/data/pageserver.toml
-        line: "{{ item }}"
-      loop:
-        - "[remote_storage]"
-        - "bucket_name = '{{ bucket_name }}'"
-        - "bucket_region = '{{ bucket_region }}'"
-        - "prefix_in_bucket = '{{ inventory_hostname }}'"
+    - name: read the existing remote pageserver config
+      ansible.builtin.slurp:
+        src: /storage/pageserver/data/pageserver.toml
+      register: _remote_ps_config
+      tags:
+      - pageserver
+
+    - name: parse the existing pageserver configuration
+      ansible.builtin.set_fact:
+        _existing_ps_config: "{{ _remote_ps_config['content'] | b64decode | sivel.toiletwater.from_toml }}"
+      tags:
+      - pageserver
+
+    - name: construct the final pageserver configuration dict
+      ansible.builtin.set_fact:
+        pageserver_config: "{{ pageserver_config_stub | combine({'id': _existing_ps_config.id }) }}"
+      tags:
+      - pageserver
+
+    - name: template the pageserver config
+      template:
+        src: templates/pageserver.toml.j2
+        dest: /storage/pageserver/data/pageserver.toml
       become: true
       tags:
       - pageserver

.github/ansible/neon-stress.hosts

@@ -1,20 +0,0 @@
-[pageservers]
-neon-stress-ps-1 console_region_id=1
-neon-stress-ps-2 console_region_id=1
-
-[safekeepers]
-neon-stress-sk-1 console_region_id=1
-neon-stress-sk-2 console_region_id=1
-neon-stress-sk-3 console_region_id=1
-
-[storage:children]
-pageservers
-safekeepers
-
-[storage:vars]
-env_name = neon-stress
-console_mgmt_base_url = http://neon-stress-console.local
-bucket_name           = neon-storage-ireland
-bucket_region         = eu-west-1
-etcd_endpoints        = neon-stress-etcd.local:2379
-safekeeper_enable_s3_offload = false

.github/ansible/neon-stress.hosts.yaml

@@ -0,0 +1,30 @@
+storage:
+  vars:
+    bucket_name: neon-storage-ireland
+    bucket_region: eu-west-1
+    console_mgmt_base_url: http://neon-stress-console.local
+    env_name: neon-stress
+    etcd_endpoints: neon-stress-etcd.local:2379
+    safekeeper_enable_s3_offload: 'false'
+    pageserver_config_stub:
+      pg_distrib_dir: /usr/local
+      remote_storage:
+        bucket_name: "{{ bucket_name }}"
+        bucket_region: "{{ bucket_region }}"
+        prefix_in_bucket: "{{ inventory_hostname }}"
+
+  children:
+    pageservers:
+      hosts:
+        neon-stress-ps-1:
+          console_region_id: 1
+        neon-stress-ps-2:
+          console_region_id: 1
+    safekeepers:
+      hosts:
+        neon-stress-sk-1:
+          console_region_id: 1
+        neon-stress-sk-2:
+          console_region_id: 1
+        neon-stress-sk-3:
+          console_region_id: 1

.github/ansible/production.hosts

@@ -1,20 +0,0 @@
-[pageservers]
-#zenith-1-ps-1 console_region_id=1
-zenith-1-ps-2 console_region_id=1
-zenith-1-ps-3 console_region_id=1
-
-[safekeepers]
-zenith-1-sk-1 console_region_id=1
-zenith-1-sk-2 console_region_id=1
-zenith-1-sk-3 console_region_id=1
-
-[storage:children]
-pageservers
-safekeepers
-
-[storage:vars]
-env_name = prod-1
-console_mgmt_base_url = http://console-release.local
-bucket_name           = zenith-storage-oregon
-bucket_region         = us-west-2
-etcd_endpoints        = zenith-1-etcd.local:2379

.github/ansible/production.hosts.yaml

@@ -0,0 +1,31 @@
+---
+storage:
+  vars:
+    env_name: prod-1
+    console_mgmt_base_url: http://console-release.local
+    bucket_name: zenith-storage-oregon
+    bucket_region: us-west-2
+    etcd_endpoints: zenith-1-etcd.local:2379
+    pageserver_config_stub:
+      pg_distrib_dir: /usr/local
+      remote_storage:
+        bucket_name: "{{ bucket_name }}"
+        bucket_region: "{{ bucket_region }}"
+        prefix_in_bucket: "{{ inventory_hostname }}"
+
+  children:
+    pageservers:
+      hosts:
+        zenith-1-ps-2:
+          console_region_id: 1
+        zenith-1-ps-3:
+          console_region_id: 1
+
+    safekeepers:
+      hosts:
+        zenith-1-sk-1:
+          console_region_id: 1
+        zenith-1-sk-2:
+          console_region_id: 1
+        zenith-1-sk-3:
+          console_region_id: 1

.github/ansible/staging.hosts

@@ -1,25 +0,0 @@
-[pageservers]
-#zenith-us-stage-ps-1 console_region_id=27
-zenith-us-stage-ps-2 console_region_id=27
-zenith-us-stage-ps-3 console_region_id=27
-zenith-us-stage-ps-4 console_region_id=27
-zenith-us-stage-test-ps-1 console_region_id=28
-
-[safekeepers]
-zenith-us-stage-sk-4 console_region_id=27
-zenith-us-stage-sk-5 console_region_id=27
-zenith-us-stage-sk-6 console_region_id=27
-zenith-us-stage-test-sk-1 console_region_id=28
-zenith-us-stage-test-sk-2 console_region_id=28
-zenith-us-stage-test-sk-3 console_region_id=28
-
-[storage:children]
-pageservers
-safekeepers
-
-[storage:vars]
-env_name = us-stage
-console_mgmt_base_url = http://console-staging.local
-bucket_name           = zenith-staging-storage-us-east-1
-bucket_region         = us-east-1
-etcd_endpoints        = zenith-us-stage-etcd.local:2379

.github/ansible/staging.hosts.yaml

@@ -0,0 +1,40 @@
+storage:
+  vars:
+    bucket_name: zenith-staging-storage-us-east-1
+    bucket_region: us-east-1
+    console_mgmt_base_url: http://console-staging.local
+    env_name: us-stage
+    etcd_endpoints: zenith-us-stage-etcd.local:2379
+    pageserver_config_stub:
+      pg_distrib_dir: /usr/local
+      remote_storage:
+        bucket_name: "{{ bucket_name }}"
+        bucket_region: "{{ bucket_region }}"
+        prefix_in_bucket: "{{ inventory_hostname }}"
+
+  children:
+    pageservers:
+      hosts:
+        zenith-us-stage-ps-2:
+          console_region_id: 27
+        zenith-us-stage-ps-3:
+          console_region_id: 27
+        zenith-us-stage-ps-4:
+          console_region_id: 27
+        zenith-us-stage-test-ps-1:
+          console_region_id: 28
+
+    safekeepers:
+      hosts:
+        zenith-us-stage-sk-4:
+          console_region_id: 27
+        zenith-us-stage-sk-5:
+          console_region_id: 27
+        zenith-us-stage-sk-6:
+          console_region_id: 27
+        zenith-us-stage-test-sk-1:
+          console_region_id: 28
+        zenith-us-stage-test-sk-2:
+          console_region_id: 28
+        zenith-us-stage-test-sk-3:
+          console_region_id: 28

.github/ansible/templates/pageserver.toml.j2

@@ -0,0 +1 @@
+{{ pageserver_config | sivel.toiletwater.to_toml }}

.github/workflows/build_and_test.yml

@@ -712,7 +712,7 @@ jobs:
       - name: Setup ansible
         run: |
           export PATH="/root/.local/bin:$PATH"
-          pip install --progress-bar off --user ansible boto3
+          pip install --progress-bar off --user ansible boto3 toml
 
       - name: Redeploy
         run: |
@@ -734,8 +734,8 @@ jobs:
           chmod 0600 ssh-key
           ssh-add ssh-key
           rm -f ssh-key ssh-key-cert.pub
-
-          ansible-playbook deploy.yaml -i ${{ matrix.env_name }}.hosts
+          ansible-galaxy collection install sivel.toiletwater
+          ansible-playbook deploy.yaml -i ${{ matrix.env_name }}.hosts.yaml
           rm -f neon_install.tar.gz .neon_current_version
 
   deploy-proxy: