Use a crontab job to keep the external data up to date

RE mantidproject#99

Linux/external-data-mirror/ansible/roles/mirror/tasks/main.yml

@@ -4,8 +4,25 @@
     state: directory
     mode: '0755'
 
+- name: Check if machine has SSH access to the main data store.
+  ansible.builtin.command: ssh {{ main_server_user }}@{{ main_server_ip }} 'echo success'
+  register: connected
+
 - name: Exchange SSH keys with linode so we can access the data.
   import_tasks: exchange-keys.yml
+  when: connected.stdout != "success"
+
+- name: Mirror the external data from the main server in a volume (this may take a while).
+  ansible.builtin.command: "rsync -az --perms -o -g {{ main_server_ip }}:/srv/{{ main_server_ip }}/ftp/external-data/MD5/ /external-data/MD5/"
+
+- name: Copy the data update script onto the mirror machine.
+  ansible.builtin.copy:
+    src: ./update-external-data.sh
+    dest: /external-data/update-external-data.sh
+    mode: '0755'
 
-- name: Mirror the external data from the main server (this may take a while).
-  command: "rsync -az --perms -o -g {{ main_server_ip }}:/srv/{{ main_server_ip }}/ftp/external-data/MD5/ /external-data/MD5/"
+- name: Create a crontab job that runs periodically to keep the data up to date.
+  ansible.builtin.cron:
+    name: Update external data
+    minute: "*/5"
+    job: /external-data/update-external-data.sh {{ main_server_ip }} >> /external-data/update-log.txt 2>&1

Linux/external-data-mirror/ansible/roles/mirror/tasks/update-external-data.sh

@@ -0,0 +1,11 @@
+#! /bin/sh
+
+SERVER_IP=${1}
+
+RSYNC_PROCESS_IDS=$(pidof rsync)
+
+if [[ -z $RSYNC_PROCESS_IDS ]]; then
+        rsync -az --perms -o -g  $SERVER_IP:/srv/$SERVER_IP/ftp/external-data/MD5/ /external-data/MD5/
+else
+        echo "rsync is already running. Skipping this time..."
+fi