WIP: Add maintenance node playbook and vars file

hosts

@@ -56,6 +56,9 @@ plausible.galaxyproject.eu
 [celerycluster]
 celery-0.galaxyproject.eu
 
+[maintenance]
+maintenance.galaxyproject.eu ansible_ssh_user=rocky
+
 # Baremetal
 [galaxyservers]
 sn06.galaxyproject.eu

maintenance.yml

@@ -0,0 +1,178 @@
+---
+- name: UseGalaxy EU maintenance server
+  hosts: maintenance
+  become: true
+  become_user: root
+  vars:
+    # The full internal name.
+    hostname: maintenance.galaxyproject.eu
+  vars_files:
+    - group_vars/gxconfig.yml # The base galaxy configuration
+    - group_vars/toolbox.yml # User controlled toolbox
+    - secret_group_vars/sentry.yml # Sentry SDK init url
+    - secret_group_vars/aws.yml # AWS creds
+    - secret_group_vars/pulsar.yml # Pulsar + MQ Connections
+    - secret_group_vars/oidc.yml # OIDC credentials (ELIXIR, keycloak)
+    - secret_group_vars/object_store.yml # Object Store credentils (S3 etc ...)
+    - secret_group_vars/db-main.yml # DB URL + some postgres stuff
+    - secret_group_vars/file_sources.yml # file_sources_conf.yml creds
+    - secret_group_vars/all.yml # All of the other assorted secrets...
+    - secret_group_vars/keys.yml # SSH keys
+    - templates/galaxy/config/job_conf.yml
+  handlers:
+    # Though this handler doesn't do anything (for now), galaxyproject.galaxy role
+    # will fail if it's not defined
+    - name: Restart Galaxy
+      shell: |
+        echo 'Manual web handler restart required' && cd /opt/galaxy/ && source /opt/galaxy/.bashrc
+  collections:
+    - devsec.hardening
+  pre_tasks:
+    - name: Install Dependencies
+      package:
+        name:
+          [
+            "git",
+            "postgresql",
+            "python3-psycopg2",
+            "python3-virtualenv",
+            "bc",
+            "python3",
+            "python3-devel",
+          ]
+      become: true
+  post_tasks:
+    - name: Append some users to the systemd-journal group
+      user:
+        name: "{{ item }}"
+        groups: systemd-journal
+        append: true
+      loop:
+        - "{{ galaxy_user.name }}"
+        - "telegraf"
+        - "stats" # special account to retrieve statistics from the server in read-only mode
+    - name: Set authorized SSH key (galaxy user)
+      ansible.posix.authorized_key:
+        user: "{{ galaxy_user.name }}"
+        state: present
+        key: "{{ item }}"
+      loop:
+        - "ecdsa-sha2-nistp256 AAAAE2VjZHNhLXNoYTItbmlzdHAyNTYAAAAIbmlzdHAyNTYAAABBBOBINXdjILF6x3WuppXyq6J2a2oSLR6waZ6txgjYJogHdIKPbI0TdReCv4EVxxYRY/NqGpHbjkqfRTsf2VgoU3U= mk@galaxy-mira"
+        - "ecdsa-sha2-nistp521 AAAAE2VjZHNhLXNoYTItbmlzdHA1MjEAAAAIbmlzdHA1MjEAAACFBACB5Q5blymkTIRSzVzXITOGvBuI7W0L9Ykwfz8LJGPraaGVPiezzFGvjhqwX+EyCqQPt7JprR5mimJRw/JN3nBXWAHjekvmB5FuILkk6m5fOiQJ5QhRMyQ5GfxODAvGbHpTuWHbYJLWD5fhcboKPxlXOWy4xY9kDZVuQvEKisNKYBsFLA== sanjay"
+    - name: Set authorized SSH key (stats user)
+      ansible.posix.authorized_key:
+        user: "stats"
+        state: present
+        key: "{{ item }}"
+      loop:
+        - "ecdsa-sha2-nistp521 AAAAE2VjZHNhLXNoYTItbmlzdHA1MjEAAAAIbmlzdHA1MjEAAACFBAA6oD5Ps9h6pKokzaAcCI6R08CKm2aSVv86h/O2HTEQkzeJq/Uvu4gbrONAM0FK5R693mHggIwaROKf1Z1+q4YNtACtYjV1c+6a9lcrGMM31y5RzO6mAW+rHNEXgZ1n3wqDcBQhSXhSekSen0R2QKwvmB7xeP4XX9qE10azZuafIFU9hQ== sanjay"
+        - "ecdsa-sha2-nistp521 AAAAE2VjZHNhLXNoYTItbmlzdHA1MjEAAAAIbmlzdHA1MjEAAACFBADKblzzPBc3+dEfFvhJQHsHGkFFN6ORjfXo71P1OutwcKEMCIcNkZKJHhYkLLrfTDN5JJ5tK2L5AaSxdwETofwm4AG1xv3LuoYsXC6e3sjKi09BVmzef520pIMW+rvL+hESwSazZaUAC0wDcH4aNDTonZYcAY87rpMX7pNMkNPJvWilUA== mira"
+  roles:
+    ## Starting configuration of the operating system
+    - role: usegalaxy_eu.handy.os_setup
+      vars:
+        enable_hostname: true
+        enable_powertools: true # geerlingguy.repo-epel role doesn't enable PowerTools repository
+        enable_remap_user: true
+        enable_exclude_packages: true
+        enable_pam_limits: true # Prevent out of control processes
+        enable_install_software: true # Some extra admin tools (*top, vim, etc)
+    - geerlingguy.repo-epel # Install EPEL repository
+
+    ## Istall miniconda, create a _galaxy_ environment and install Packages
+    ## Galaxy will use the virtualenv command from this conda environment (see
+    ## galaxy_virtualenv_command) in the group_vars/maintenance.yml
+    - role: galaxyproject.miniconda
+      vars:
+        miniconda_prefix: "{{ conda_prefix }}"
+        galaxy_conda_create_env: true
+        galaxy_conda_env_packages:
+          - python=3.8.8
+          - pip
+          - virtualenv
+
+    - usegalaxy-eu.autoupdates # keep all of our packages up to date
+    - influxdata.chrony # Keep our time in sync.
+
+    ## Filesystems
+    - usegalaxy-eu.autofs # Setup the mount points which will be needed later
+
+    ## Monitoring (Uncomment when in production)
+    # - hxr.monitor-cluster
+    # - hxr.monitor-email
+    # - usegalaxy-eu.monitoring
+
+    ## Setup Galaxy user
+    - role: galaxyproject.galaxy
+      vars:
+        galaxy_create_user: true
+        galaxy_manage_clone: false
+        galaxy_manage_download: false
+        galaxy_manage_existing: true
+        galaxy_manage_paths: true
+        galaxy_manage_static_setup: false
+        galaxy_manage_mutable_setup: false
+        galaxy_manage_database: false
+        galaxy_fetch_dependencies: false
+        galaxy_build_client: false
+
+    ## The bashrc needs to be created for several later features.
+    - role: usegalaxy-eu.bashrc
+
+    # HTCondor (Uncomment when in production and latest version is installed)
+    # - usegalaxy_eu.htcondor
+
+    # Misc.
+    - role: hxr.galaxy-nonreproducible-tools
+      become: true
+      become_user: galaxy
+    - usegalaxy-eu.dynmotd # nicer MOTD/welcome message
+    - usegalaxy-eu.rsync-galaxy-sync # syncs Galaxy codebase to NFS and to the headnodes
+    - role: usegalaxy-eu.webhooks # Clone webhook repository
+      become: true
+      become_user: galaxy
+    - role: usegalaxy-eu.tours # Clone tour repository
+      become: true
+      become_user: galaxy
+
+    ## SSL / Security
+    # https://en.wikipedia.org/wiki/Trust_on_first_use
+    - ssh-host-sign # Sign the server host key to prevent TOFU for SSH
+
+    ## GALAXY
+    - hxr.postgres-connection
+    - galaxyproject.gxadmin
+    # TODO move under monitoring + telegraf.
+    # Role that creates a cron to gather galaxy stats, needs to run only on one node (uncomment when in production)
+    # - usegalaxy-eu.galaxy-slurp
+
+    # The REAL galaxy role
+    - role: galaxyproject.galaxy
+      vars:
+        galaxy_create_user: true
+        galaxy_manage_clone: true
+        galaxy_manage_download: false
+        galaxy_manage_existing: false
+        galaxy_manage_static_setup: true
+        galaxy_manage_mutable_setup: true
+        galaxy_manage_database: true
+        galaxy_fetch_dependencies: true
+        galaxy_build_client: true
+
+    ## Extras!
+    - usegalaxy-eu.fix-galaxy-server-dir # Fix details into the galaxy server dirs
+    - hxr.install-to-venv # Some extra packages our site needs.
+    - usegalaxy_eu.gie_proxy # Setup the NodeJS proxy for GxIT (not IE anymore)
+    # - usegalaxy_eu.fs_maintenance # (Uncomment when in production)
+    # - usegalaxy-eu.htcondor_release # (Uncomment when in production and condor is installed)
+    # Various ugly fixes
+    # - usegalaxy-eu.fix-unscheduled-workflows (Uncomment when in production)
+    # - usegalaxy-eu.fix-ancient-ftp-data # Remove FTP data older than 3 months, create FTP user directories (Uncomment when in production)
+    # - usegalaxy-eu.fix-missing-api-keys # Workaround for IE users not have a key set. (Uncomment when in production)
+    # - usegalaxy-eu.fix-user-quotas # Automatically recalculate user quotas and attribute ELIXIR quota to ELIXIR AAI user on a regular basis (Uncomment when in production)
+    - usegalaxy_eu.tpv_auto_lint
+    - ssh_hardening #dev-sec.hardening collection
+    - dj-wasabi.telegraf    # Uncomment when in production
+    - usegalaxy-eu.logrotate # Rotate logs
+    - usegalaxy-eu.fix-stop-ITs
+    - usegalaxy-eu.vgcn-monitoring

requirements.yaml

@@ -21,7 +21,7 @@ collections:
     source: https://galaxy.ansible.com
     type: galaxy
   - name: usegalaxy_eu.handy
-    version: 2.10.0
+    version: 2.11.1
     source: https://galaxy.ansible.com
 
 roles:
@@ -89,7 +89,7 @@ roles:
     src: https://github.com/usegalaxy-eu/ansible-update-hosts
     version: 0.2.0
   - name: usegalaxy_eu.gie_proxy
-    version: 0.0.2
+    version: 0.0.4
   - name: usegalaxy-eu.autofs
     src: https://github.com/usegalaxy-eu/ansible-autofs
     version: 1.0.0

roles/dj-wasabi.telegraf/tasks/RedHat.yml

@@ -7,7 +7,7 @@
     description: InfluxDB Repository - RHEL $releasever
     baseurl: "https://repos.influxdata.com/rhel/{{ telegraf_redhat_releasever }}/$basearch/stable"
     gpgcheck: yes
-    gpgkey: https://repos.influxdata.com/influxdb.key
+    gpgkey: https://repos.influxdata.com/influxdata-archive_compat.key
 
 - name: "Install telegraf package | RedHat"
   yum:

roles/usegalaxy-eu.rsync-galaxy-sync/README.md

@@ -0,0 +1,30 @@
+Role Name
+=========
+
+Adds a rsync script that performs a full sync of a Galaxy codebase to an NFS share and the head nodes
+
+Role Variables
+--------------
+
+`execute_galaxy_sync`: Whether to execute the sync script or not. Defaults to `false`
+`galaxy_rsync_user_private_key_file`: The private key of the user that will be used to rsync the codebase. If this key does not exist then it will be added from the vault file.
+`headnodes`: A list of headnodes to rsync the codebase to. Defaults to `sn07.galaxyproject.eu` (this is currently (24/03/2023) in testing phase so the default is `sn07`)
+`headnodes_sync_location`: The location on the headnodes to rsync the codebase to. Defaults to the variable `galaxy_root` (which is defined in the group_vars files)
+
+Dependencies
+------------
+
+- `prsync` command (if not installed, it will be installed. The command is available in the `pssh` package)
+
+Example Playbook
+----------------
+
+    - hosts: maintenance
+      roles:
+         - role: usegalaxy-eu.rsync-galaxy-sync
+           vars:
+              execute_galaxy_sync: false
+              galaxy_rsync_user_private_key: "/opt/galaxy/.ssh/galaxy_rsync_key"
+              headnodes: "sn07.galaxyproject.eu"
+              headnodes_sync_location: "/opt/galaxy"
+              galaxy_nfs_location: "/data/galaxy-sync"

roles/usegalaxy-eu.rsync-galaxy-sync/defaults/main.yml

@@ -0,0 +1,9 @@
+---
+execute_galaxy_sync: false
+galaxy_rsync_user_private_key_file: "{{ galaxy_user.home }}/.ssh/galaxy_rsync_key"
+
+# Uses the hostnames of the inventory group 'galaxyservers'
+# headnodes: "{{ groups['galaxyservers'] | join(', ') }}"
+headnodes: "sn07.galaxyproject.eu"
+headnodes_sync_location: "{{ galaxy_root }}"
+galaxy_nfs_location: ""

roles/usegalaxy-eu.rsync-galaxy-sync/meta/main.yml

@@ -0,0 +1,14 @@
+---
+galaxy_info:
+  author: The Galaxy Project
+  description: Installs a Galaxy rsync script
+  company: The Galaxy Project
+  license: AFL v3.0
+  min_ansible_version: 2.5
+  platforms:
+  - name: EL
+    versions:
+    - 8
+    - 9
+  galaxy_tags: []
+dependencies: []

roles/usegalaxy-eu.rsync-galaxy-sync/tasks/main.yml

@@ -0,0 +1,80 @@
+---
+- name: Check if prsync is installed
+  command: prsync --version
+  register: prsync_installed
+  ignore_errors: true
+
+- name: Install prsync (part of pssh)
+  become: true
+  dnf:
+    name: pssh
+    state: present
+  when: not prsync_installed.stdout
+
+- name: Check if SSH key exists
+  stat:
+    path: "{{ galaxy_rsync_user_private_key_file }}"
+  register: ssh_key
+
+- name: Create SSH directory
+  file:
+    path: "{{ galaxy_user.home }}/.ssh"
+    state: directory
+    owner: "{{ galaxy_user.name }}"
+    group: "{{ galaxy_user.name }}"
+    mode: 0700
+  when: not ssh_key.stat.exists
+
+- name: Add SSH key
+  copy:
+    content: "{{ galaxy_user_private_key }}"
+    dest: "{{ galaxy_rsync_user_private_key_file }}"
+    owner: "{{ galaxy_user.name }}"
+    group: "{{ galaxy_user.name }}"
+    mode: 0600
+  when: not ssh_key.stat.exists
+
+- name: "Deploy galaxy-rsync script"
+  copy:
+    content: |
+        #!/bin/bash
+        headnodes="{{ headnodes }}"
+        cd {{ galaxy_root }};
+        for dir in {config,custom-tools,dynamic_rules,gie-proxy,mutable-config,mutable-data,server,venv,tool-data}; do
+            if [ -d $dir ]; then
+                echo "Syncing $dir"
+                # Sync to NFS server in background
+                rsync -avr --delete --exclude node_modules/ --exclude .git --exclude __pycache__ $dir/ {{ galaxy_nfs_location }}/$dir/ &
+
+                # Sync to headnodes only if the variable is set
+                if [ ! -z $headnodes ]; then
+                    # Sync to head nodes in foreground (so we can see progress and wait until the sync is done before continuing with the next directory)
+                    prsync -avr -H $headnodes --extra-arg='--delete' --extra-arg='--exclude=node_modules/' --extra-arg='--exclude=.git' --extra-arg='--exclude=__pycache__' --user "{{ galaxy_user.name }}" --ssh-args='-i "{{ galaxy_rsync_user_private_key_file }}"' $dir/ {{ headnodes_sync_location }}/$dir/
+                fi
+            else
+                echo "Skipping $dir"
+            fi
+        done;
+        if [ -d shed_tools-local ]; then
+            echo "Syncing shed_tools-local"
+            # Sync to NFS server in background
+            rsync -avr --delete --exclude .hg shed_tools-local/ {{ galaxy_nfs_location }}/shed_tools/ &
+
+            # Sync to headnodes only if the variable is set
+                if [ ! -z $headnodes ]; then
+                    # Sync to head nodes in foreground (so we can see progress and wait until the sync is done before continuing with the next directory)
+                    prsync -avr -H $headnodes --extra-arg='--delete' --extra-arg='--exclude=.hg' --user "{{ galaxy_user.name }}" --ssh-args='-i "{{ galaxy_rsync_user_private_key_file }}"' shed_tools-local/ {{ headnodes_sync_location }}/shed_tools/
+                fi
+        else
+            echo "Skipping shed_tools-local"
+        fi
+    dest: /usr/bin/galaxy-rsync
+    owner: root
+    group: root
+    mode: 0755
+
+- name: "Execute the script. Syncing in progress."
+  command: /usr/bin/galaxy-rsync
+  become: true
+  become_user: galaxy
+  when: execute_galaxy_sync