Re: [pve-devel] [PATCH qemu-server 3/3] api: include not mapped resources for running vms in migrate preconditions

[Fiona Ebner <f.ebner@proxmox.com>]

Am 02.04.24 um 11:39 schrieb Dominik Csapak:
> On 3/22/24 17:19, Fiona Ebner wrote:
>> Am 20.03.24 um 13:51 schrieb Dominik Csapak:
>>> so that we can show a proper warning in the migrate dialog and check it
>>> in the bulk migrate precondition check
>>>
>>> the unavailable_storages and allowed_nodes should be the same as before
>>>
>>> Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
>>> ---
>>> not super happy with this partial approach, we probably should just
>>> always return the 'allowed_nodes' and 'not_allowed_nodes' and change
>>> the gui to handle the running vs not running state?
>>
>> So not_allowed_nodes can already be returned in both states after this
>> patch. But allowed nodes still only if not running. I mean, there could
>> be API users that break if we'd always return allowed_nodes, but it
>> doesn't sound unreasonable for me to do so. Might even be an opportunity
>> to structure the code in a bit more straightforward manner.
> 
> yes, as said previosly i'd like this api call a bit to make it more
> practical
> but that probably has to wait for the next major release
> 
> as for returning 'allowed_nodes' always, we'd have to adapt the gui of
> course,
> but if we don't deem it 'too breaking' i'd rework that a bit even now
> 

Thinking about it in general for existing API users:

1. If allowed_nodes is not checked for live-migration, no breakage.

2. If allowed_nodes is checked for live-migration, the API user just
becomes more accurate (as long as what we return is correct).

3. If there is an assert that allowed_nodes is not returned for
live-migration, breakage.

4. If presence of allowed_nodes is used to guess whether it's a
live-migration or not, breakage. But this is just a bug IMHO.