Storage removal

[pkliczewski]

Datastores can't be removed if there are relations to hosts or vms.
We need provide a way to check whether datastore removal is supported
(no vm or host relations). When there condition is not met we need
to notify a user that relations needs to be removed first.

This PR fixes:
https://bugzilla.redhat.com/1439380

[pkliczewski]

This PR depends on ManageIQ/manageiq#14724

[pkliczewski]

Travis needs ManageIQ/manageiq#14724 first and the test will pass.

[pkliczewski]

@miq-bot assign @martinpovolny

[pkliczewski]

@borod108 @masayag please take a look

[miq-bot]

Checked commit pkliczewski@2cfe2cf with ruby 2.2.6, rubocop 0.47.1, and haml-lint 0.20.0
1 file checked, 0 offenses detected
Everything looks good. 🍰

[pkliczewski]

Bug 1439380 - After the removal of a rhev provider the datastores are still displayed but unable to be deleted

[martinpovolny]

Can this to be coded using the supports mechanism on the backend? (See e.g. app/models/vm_or_template/operations/snapshot.rb in the core repo).

And then I would expect the support to be checked in the code that displayes the button and the error message used from the support feature.

Then here we could check again (and again reuse the error message).

Does it make sense?

[pkliczewski]

It is dependent on backend and it is using supports. Please see ManageIQ/manageiq#14724

[martinpovolny]

Right, so then we should reuse the "unsuported reason" to form the flash message and it's done.

Something like:

add_flash(_(SecurityGroup.unsupported_reason(:create)), :error)

but I don't know how that's done with batch_operation_supported?

[pkliczewski]

OK, will fix

[durandom]

why is this fine/yes?
https://bugzilla.redhat.com/1439380 is not?!

[martinpovolny]

why is this fine/yes?
https://bugzilla.redhat.com/1439380 is not?!

@dclarizio : shall this go to fine?

[blomquisg]

@martinpovolny @durandom I've gotten clarification from PM that this can go to Fine. However, there's some discussion on the related Backend PR as to the general solution. Need to make sure we're all on the same page before moving forward.

[martinpovolny]

@pkliczewski : I merged this as the backend is merged. Please, deal with the error message in a follow up PR (and ping me). Thx!

[pkliczewski]

@martinpovolny Sure, I will.

[durandom]

@miq-bot add_label fine/yes
because BZ got updated

[durandom]

@miq-bot remove_label fine/no

[simaishi]

Fine backport details:

 $ git log -1
commit 11807c93192667e8745655155aa46450434bbdaa
Author: Martin Povolny <[email protected]>
Date:   Mon Apr 24 07:15:33 2017 +0200

    Merge pull request #993 from pkliczewski/master
    
    Storage removal
    (cherry picked from commit befd4da221979024c715812dd3ad5c5a75ff4425)
    
    https://bugzilla.redhat.com/show_bug.cgi?id=1444889

[pkliczewski]

@martinpovolny I checked and it is not that easy to get unsupported_reason for a group of storage domains. We can have 1 or more domains. We would need to decide which one to display. It seems like this approach (having generic message) fits good in such use case.

[martinpovolny]

@pkliczewski : ok, thx

[durandom]

@pkliczewski could you add this shortcoming to ManageIQ/manageiq#14740 ? I'll collect there what is needed as improvemnts in the supports feature mixin.

[pkliczewski]

@durandom Done