Fix issue with Swift Recon middleware

Follow-up on I343d8f45a78ebc3c11ed0c68fe8bec24f9ea7929

According the documentation [1] we forgot to share statistics data
for swift-account-server and swift-container-server. This change will
fix the issue.

1. https://docs.openstack.org/swift/latest/admin_guide.html#cluster-telemetry-and-monitoring

Closes-Bug: #1941611
Change-Id: Ib9afd84cac1fcbd96f98b4720ea9c6503bbdb124
Signed-off-by: Maksim Malchuk <[email protected]>

ansible/roles/swift/tasks/start.yml

@@ -60,6 +60,7 @@
       - "{{ node_config_directory }}/swift-account-server/:{{ container_config_directory }}/:ro"
       - "{{ swift_devices_mount_point }}:{{ swift_devices_mount_point }}:shared"
       - "/etc/localtime:/etc/localtime:ro"
+      - "swift_recon_cache:/var/cache/swift"
   when: inventory_hostname in groups['swift-account-server']
 
 - name: Starting swift-account-auditor container
@@ -138,6 +139,7 @@
       - "{{ node_config_directory }}/swift-container-server/:{{ container_config_directory }}/:ro"
       - "{{ swift_devices_mount_point }}:{{ swift_devices_mount_point }}:shared"
       - "/etc/localtime:/etc/localtime:ro"
+      - "swift_recon_cache:/var/cache/swift"
   when: inventory_hostname in groups['swift-container-server']
 
 - name: Starting swift-container-auditor container