[tripleo_transfer] Do transfers directly from src to dst

When running fetch with become, the slurp module will also
be used to fetch the contents of the file for determining
the remote checksum. This effectively doubles the transfer
size [0] and shows up as a MemoryError when the file size
is large enough.

In TripleO this is problematic in large & old deployments
when transferring the /var/lib/mysql folder.

This patch switches to using rsync directly between the src
and dst hosts to transfer the data. This is advantageous not
only for solving the above-mentioned bug, but is also faster.

A simpler implementation using synchronize was attempted [1],
but there were issues with the mistral container which
prevented that approach from being successful.

[0] https://docs.ansible.com/ansible/latest/collections/ansible/builtin/fetch_module.html#notes
[1] https://review.opendev.org/c/openstack/tripleo-ansible/+/776565/11

Closes-Bug: #1908425
Closes-Bug: rhbz#1904681
Closes-Bug: rhbz#1916162
Depends-On: https://review.opendev.org/778650

Change-Id: Ifc03f9eb1cb4ca3faec194569f4cb2dace93323f
(cherry picked from commit d7730980c9)

tripleo_ansible/roles/tripleo-transfer/defaults/main.yml

@@ -23,10 +23,12 @@
 #   * `tripleo_transfer_src_dir` -- directory on the source host to transfer from
 #   * `tripleo_transfer_dest_host` -- the inventory name of the destination host
 #   * `tripleo_transfer_dest_dir` -- directory on the destination host to transfer to
-
-tripleo_transfer_storage_root_dir: /var/lib/mistral/tripleo-transfer
-tripleo_transfer_storage_root_become: false
 tripleo_transfer_src_become: true
 tripleo_transfer_dest_become: true
-tripleo_transfer_dest_wipe: true
 tripleo_transfer_flag_file: ~
+# tripleo_transfer_key_location: location of the private key used to connect
+# from src host to dest host.
+tripleo_transfer_key_location: "~/transfer_key"
+# tripleo_transfer_cleanup_keys: clean up the keypair from the source host
+# and remove public key from destination host when true.
+tripleo_transfer_cleanup_keys: true

tripleo_ansible/roles/tripleo-transfer/molecule/default/playbook.yml

@@ -14,6 +14,14 @@
 # License for the specific language governing permissions and limitations
 # under the License.
 
+- name: Collect facts
+  hosts: all
+  gather_facts: false
+  any_errors_fatal: true
+  tasks:
+    - name: Gather a minimal set of facts
+      setup:
+        gather_subset: '!all,min'
 
 - name: Converge
   hosts: localhost
@@ -25,4 +33,3 @@
       tripleo_transfer_src_dir: /etc
       tripleo_transfer_dest_host: controller1
       tripleo_transfer_dest_dir: /opt/etc-target
-      tripleo_transfer_storage_root_dir: /tmp/transfer-staging

tripleo_ansible/roles/tripleo-transfer/tasks/flag.yml

@@ -19,8 +19,6 @@
   stat:
     path: "{{ tripleo_transfer_flag_file }}"
   register: tripleo_transfer_flag_stat
-  become: "{{ tripleo_transfer_dest_become }}"
-  delegate_to: "{{ tripleo_transfer_dest_host }}"
 
 - name: fail if flag file exists
   fail:
@@ -32,18 +30,13 @@
       and re-run the data transfer.
   when:
     - tripleo_transfer_flag_stat.stat.exists
-  delegate_to: "{{ tripleo_transfer_dest_host }}"
 
 - name: ensure directory for flag file exists
   file:
     path: "{{ tripleo_transfer_flag_file|dirname }}"
     state: directory
-  become: "{{ tripleo_transfer_dest_become }}"
-  delegate_to: "{{ tripleo_transfer_dest_host }}"
 
 - name: create the flag file
   file:
     path: "{{ tripleo_transfer_flag_file }}"
     state: touch
-  become: "{{ tripleo_transfer_dest_become }}"
-  delegate_to: "{{ tripleo_transfer_dest_host }}"

tripleo_ansible/roles/tripleo-transfer/tasks/main.yml

@@ -15,89 +15,87 @@
 # under the License.
 
 
-- name: make sure we don't have a trailing forward slash in the src
+# Note:
-  set_fact:
+# This role is used in a playbook that typically targets the undercloud
-    tripleo_transfer_src_dir_safe: "{{ tripleo_transfer_src_dir | regex_replace('\\/$', '') }}"
+# and may target other hosts, so to ensure that it executes against the
-    cacheable: false
+# right src and dest it uses delegation throughout.
 
-- name: make sure we don't have a trailing forward slash in the dst
+- name: tripleo_transfer tasks
-  set_fact:
+  run_once: true
-    tripleo_transfer_dest_dir_safe: "{{ tripleo_transfer_dest_dir | regex_replace('\\/$', '') }}"
+  block:
-    cacheable: false
+    - name: install requirements in src and dest hosts
+      become: true
+      package:
+        name:
+          - rsync
+          - openssh-clients
+        state: present
+      delegate_to: "{{ item }}"
+      loop:
+        - "{{ tripleo_transfer_src_host }}"
+        - "{{ tripleo_transfer_dest_host }}"
 
-- name: ensure local storage directory exists and has correct permissions
+    - name: generate ssh key-pair in source host
-  file:
+      openssh_keypair:
-    path: "{{ tripleo_transfer_storage_root_dir }}"
+        path: "{{ tripleo_transfer_key_location }}"
-    # Attempting to set an owner fails with "chown failed: failed to
+      delegate_to: "{{ tripleo_transfer_src_host }}"
-    # look up user" so we at least ensure the permissions.
+      become: "{{ tripleo_transfer_src_become }}"
-    mode: 0700
+      register: keypair_gen
-    state: directory
-  delegate_to: localhost
-  become: "{{ tripleo_transfer_storage_root_become }}"
 
-- name: create tempfile for the archive
+    - name: set authorized-keys in destination host
-  tempfile:
+      authorized_key:
-    prefix: ansible.tripleo-transfer.
+        comment: "Added by tripleo-transfer"
-  register: tripleo_transfer_tempfile
+        user: "{{ ansible_user|default(ansible_ssh_user|default(hostvars[tripleo_transfer_dest_host].ansible_user_id)) }}"
-  become: "{{ tripleo_transfer_src_become }}"
+        state: present
-  delegate_to: "{{ tripleo_transfer_src_host }}"
+        key: "{{ keypair_gen.public_key }}"
+      delegate_to: "{{ tripleo_transfer_dest_host }}"
+      when: keypair_gen is succeeded
 
-# Using the "archive" module lists lists all tarred files in module
+    - import_tasks: flag.yml
-# output, if there's too many files, it can crash ansible even with
+      when:
-# "no_log: true".
+        - tripleo_transfer_flag_file != None
-- name: create the archive
+        - tripleo_transfer_flag_file|length > 0
-  shell: |-
+      become: "{{ tripleo_transfer_dest_become }}"
-    set -euo pipefail
+      delegate_to: "{{ tripleo_transfer_dest_host }}"
-    tar --transform "s|^{{ tripleo_transfer_src_dir_safe | basename }}|{{ tripleo_transfer_dest_dir_safe | basename }}|" \
-        -czf "{{ tripleo_transfer_tempfile.path }}" \
-        -C "{{ tripleo_transfer_src_dir_safe | dirname }}" \
-        "{{ tripleo_transfer_src_dir_safe | basename }}"
-  become: "{{ tripleo_transfer_src_become }}"
-  delegate_to: "{{ tripleo_transfer_src_host }}"
 
-- name: fetch the archive
+    - name: synchronize both directories
-  fetch:
+      vars:
-    src: "{{ tripleo_transfer_tempfile.path }}"
+        hostvars_dest_host_ip: >-
-    dest: "{{ tripleo_transfer_storage_root_dir }}/{{ tripleo_transfer_dest_host }}{{ tripleo_transfer_dest_dir_safe }}.tar.gz"
+          {{ hostvars[tripleo_transfer_dest_host].ansible_host |
-    flat: true
+          default(hostvars[tripleo_transfer_dest_host].inventory_hostname) }}
-  become: "{{ tripleo_transfer_src_become }}"
+        tripleo_transfer_dest_user: >-
-  delegate_to: "{{ tripleo_transfer_src_host }}"
+          {{ hostvars[tripleo_transfer_dest_host].ansible_user |
+          default(hostvars[tripleo_transfer_dest_host].ansible_ssh_user |
+          default(hostvars[tripleo_transfer_dest_host].ansible_user_id)) }}
+      shell: >-
+        /usr/bin/rsync
+        -v
+        --delay-updates
+        -F
+        --compress
+        --archive
+        --delete
+        --rsync-path='sudo rsync'
+        --rsh='ssh -o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null -i {{ tripleo_transfer_key_location }}'
+        {{ tripleo_transfer_src_dir_safe }}
+        {{ tripleo_transfer_dest_user }}@{{ hostvars_dest_host_ip }}:{{ tripleo_transfer_dest_dir_safe }}
+      become: "{{ tripleo_transfer_src_become }}"
+      delegate_to: "{{ tripleo_transfer_src_host }}"
+  always:
+    - name: clean up keys in source host
+      file:
+        path: "{{ item }}"
+        state: absent
+      delegate_to: "{{ tripleo_transfer_src_host }}"
+      become: "{{ tripleo_transfer_src_become }}"
+      loop:
+        - "{{ tripleo_transfer_key_location }}"
+        - "{{ tripleo_transfer_key_location }}.pub"
 
-- name: remove tempfile
+    - name: remove public key from authorized keys in destination host
-  file:
+      lineinfile:
-    name: "{{ tripleo_transfer_tempfile.path }}"
+        path: "~/.ssh/authorized_keys"
-    state: absent
+        state: absent
-  become: "{{ tripleo_transfer_src_become }}"
+        regexp: '.*Added by tripleo-transfer.*$'
-  delegate_to: "{{ tripleo_transfer_src_host }}"
+      delegate_to: "{{ tripleo_transfer_dest_host }}"
-
+  when: tripleo_transfer_cleanup_keys | bool
-- include_tasks: flag.yml
-  when:
-    - tripleo_transfer_flag_file != None
-    - tripleo_transfer_flag_file|length > 0
-
-- name: wipe the destination directory
-  file:
-    path: "{{ tripleo_transfer_dest_dir_safe }}"
-    state: absent
-  become: "{{ tripleo_transfer_dest_become }}"
-  delegate_to: "{{ tripleo_transfer_dest_host }}"
-  when: tripleo_transfer_dest_wipe|bool
-
-- name: make sure the destination parent directory is present
-  file:
-    path: "{{ tripleo_transfer_dest_dir_safe|dirname }}"
-    state: directory
-  become: "{{ tripleo_transfer_dest_become }}"
-  delegate_to: "{{ tripleo_transfer_dest_host }}"
-
-- name: push and extract the archive
-  unarchive:
-    src: "{{ tripleo_transfer_storage_root_dir }}/{{ tripleo_transfer_dest_host }}{{ tripleo_transfer_dest_dir_safe }}.tar.gz"
-    dest: "{{ tripleo_transfer_dest_dir_safe|dirname }}"
-  become: "{{ tripleo_transfer_dest_become }}"
-  delegate_to: "{{ tripleo_transfer_dest_host }}"
-
-- name: remove the local archive
-  file:
-    path: "{{ tripleo_transfer_storage_root_dir }}/{{ tripleo_transfer_dest_host }}{{ tripleo_transfer_dest_dir_safe }}.tar.gz"
-    state: absent

tripleo_ansible/roles/tripleo-transfer/vars/main.yml

@@ -15,9 +15,11 @@
 # under the License.
 
 
-- name: ensure tripleo_transfer storage directory is removed
+# All variables intended for internal use should be placed in this file.
-  file:
+
-    path: "{{ tripleo_transfer_storage_root_dir }}"
+
-    state: absent
+# make sure we have a trailing forward slash in the src, otherwise rsync creates extra dir
-  delegate_to: localhost
+tripleo_transfer_src_dir_safe: "{{ tripleo_transfer_src_dir }}/"
-  become: "{{ tripleo_transfer_storage_root_become }}"
+
+# make sure we do not have a trailing forward slash in the dest
+tripleo_transfer_dest_dir_safe: "{{ tripleo_transfer_dest_dir | regex_replace('\\/$', '') }}"