Note
Before proceeding with the manual steps below, verify that you have performed the steps described in Apply maintenance updates.
Fixed the issue with the apache2 service failing to start after rebooting
of any proxy node. The issue occurred due to Apache requiring the certificates
placed on the GlusterFS volume, which might not have been mounted before the
apache2 service start. To apply the issue resolution, set the dependency
between the GlusterFS volume mount and the apache2 service explicitly as
described below.
To apply the issue resolution:
Log in to the Salt Master node.
In classes/cluster/<cluster_name>/openstack/proxy.yml,
add the following pillar:
parameters:
...
apache:
server:
wait_for_service:
- srv-salt-pki.mount
...
Apply the changes to Apache on the proxy nodes:
salt 'prx*' saltutil.refresh_pillar
salt 'prx*' state.apply apache.server
Fixed the issue with the Sphinx providing Reclass variables without
authorization. The issue occurred due to NGINX being configured
without a basic authorization for reclass_doc.
To apply the issue resolution:
Log in to the Salt Master node.
Generate a Sphinx UI password and add it to
classes/cluster/<cluster_name>/infra/secrets.yml using the following
parameter as described in
MCP Operations Guide: Manage secrets in the Reclass model.
parameters:
_param:
...
sphinx_proxy_password_generated: <generated_password>
...
Apply the changes to NGINX and Sphinx:
salt -C 'I@sphinx:server and I@nginx:server' saltutil.refresh_pillar
salt -C 'I@sphinx:server and I@nginx:server' state.apply nginx
Fixed the issue with a wrong default value set for umask, which caused the
Cannot access storage file: Permission denied exception when running the
salt.control state if umask was set to 027.
To apply the issue resolution:
Log in to the Salt Master node.
Apply the following state:
salt -C "I@salt:control" state.apply salt.control