Misleading asynchronous user message when share services are down

Description
===========
When all of the share manager services are down, or are still initializing, the scheduler has no information to make an informed decision for placement. When this happens, the user message one receives is quite confusing, and should be enhanced.

Steps to reproduce
==================

A chronological list of steps which will help reproduce the issue you hit:
* Configure manila with a share driver that fails to initialize. You can do this by misconfiguring your share backend such that the driver cannot connect to the storage system, and keeps retrying.
* Create a share, the share status will transition from "creating" to "error".
* Check user messages ('manila message-list')

Expected result
===============
The user message should suggest that there is a problem with the environment and isn't ready for provisioning yet

Actual result
=============
The user message is as follows:

$ manila message-list --resource-id b0283ea1-6a40-484a-bc5c-fea25df6f3cd
+--------------------------------------+---------------+--------------------------------------+-----------+--------------------------------------------------------------------------------------------------------------+-----------+----------------------------+
| ID | Resource Type | Resource ID | Action ID | User Message | Detail ID | Created At |
+--------------------------------------+---------------+--------------------------------------+-----------+--------------------------------------------------------------------------------------------------------------+-----------+----------------------------+
| ca92cb0f-da53-4c97-88a2-f0e6e6f02337 | SHARE | b0283ea1-6a40-484a-bc5c-fea25df6f3cd | 001 | allocate host: No storage could be allocated for this share request, AvailabilityZone filter didn't succeed. | 007 | 2020-07-07T17:10:11.000000 |
+--------------------------------------+---------------+--------------------------------------+-----------+--------------------------------------------------------------------------------------------------------------+-----------+----------------------------+

$ manila message-show ca92cb0f-da53-4c97-88a2-f0e6e6f02337
+---------------+--------------------------------------------------------------------------------------------------------------+
| Property | Value |
+---------------+--------------------------------------------------------------------------------------------------------------+
| id | ca92cb0f-da53-4c97-88a2-f0e6e6f02337 |
| resource_type | SHARE |
| resource_id | b0283ea1-6a40-484a-bc5c-fea25df6f3cd |
| action_id | 001 |
| user_message | allocate host: No storage could be allocated for this share request, AvailabilityZone filter didn't succeed. |
| message_level | ERROR |
| detail_id | 007 |
| created_at | 2020-07-07T17:10:11.000000 |
| expires_at | 2020-08-06T17:10:11.000000 |
| request_id | req-81c4d5f9-eba3-48bd-a04c-a1d7bfeaa97e |
+---------------+--------------------------------------------------------------------------------------------------------------+

The scheduler log file has this logging:

ubuntu-bionic-rax-iad-0017829885 manila-scheduler[664]: ERROR manila.scheduler.manager [None req-81c4d5f9-eba3-48bd-a04c-a1d7bfeaa97e None None] Failed to schedule create_share: No valid host was found. Failed to find a weighted host, the last executed filter was AvailabilityZoneFilter.: manila.exception.NoValidHost: No valid host was found. Failed to find a weighted host, the last executed filter was AvailabilityZoneFilter.

Environment
===========
1. Exact version of OpenStack Manila you are running. See the following
  list for all releases: trunk

2. Which storage backend did you use?
   Dummy driver, the situation is backend agnostic

3. Which networking type did you use?
   Neutron with OpenVSwitch, the problem occurs agnostic to the DHSS mode