Description of problem:
When you increase or decrease the number of cores per socket while VM is up, the request in CFME UI is completed successfully. However, the changed has not been actually applied since RHV does not support this. CFME user is not warned about this in any way. They should know that this change will be applied __only once the VM has been restarted__.
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1. Have a CFME appliance with RHV as a provider. Have a running VM on RHV.
2. Go to the VM's detail page.
3. Click Configuration -> Reconfigure this VM. Choose Processors switch and increase the number of Cores per Sockets.
4. Submit your configuration request and wait for it to finish.
Request is completed successfully. However, if you inspect the VM in RHV or if you SSH to it, you can see that the number of cores per socket has not increased. There is no indication in CFME UI that this is the case.
I would expect as an user to be informed that some configuration is waiting for the VM to be restarted in order to actually take place. Right now there is inconsistency. If you change VM's memory or number of sockets, you get the same result. However, unlike cores per socket, memory and sockets change takes place immediately without the need to restart the VM.
New commit detected on ManageIQ/manageiq-schema/master:
Author: Roberto Ciatti <email@example.com>
AuthorDate: Thu Nov 28 06:22:50 2019 -0500
Commit: Roberto Ciatti <firstname.lastname@example.org>
CommitDate: Thu Nov 28 06:22:50 2019 -0500
Add restart_needed field to vms
Sometimes VMs need to be restarted to apply not hotpluggable configuration
changes and the user needs to be aware of that.
Changes to a VM can be applied through ManageIQ but also by the
external the external system
This is part of the RFE
db/migrate/20191128111630_add_restart_needed_to_vms.rb | 5 +
1 file changed, 5 insertions(+)
New commit detected on ManageIQ/manageiq-ui-classic/master:
Author: Milan Zazrivec <email@example.com>
AuthorDate: Wed Jul 22 16:18:36 2020 +0000
Commit: Milan Zazrivec <firstname.lastname@example.org>
CommitDate: Thu Jul 23 11:19:23 2020 +0000
Add flash message about required restart to VM summary screen
app/controllers/vm_common.rb | 6 +
app/helpers/textual_summary_helper.rb | 6 +
app/helpers/vm_helper/textual_summary.rb | 4 +
app/views/layouts/_textual_groups_generic.html.haml | 4 +
4 files changed, 20 insertions(+)
There's no need to use automation here. You can create a custom report directly showing
the "Restart Needed" attribute (plus whatever else you need to show).
What I wrote in comment #15 was that this feature (overall, not just the UI part) requires
a schema change and as such most likely will not land in CF 5.0 / CFME 5.11.
(In reply to Milan Zázrivec from comment #18)
> There's no need to use automation here. You can create a custom report
> directly showing
> the "Restart Needed" attribute (plus whatever else you need to show).
> What I wrote in comment #15 was that this feature (overall, not just the UI
> part) requires
> a schema change and as such most likely will not land in CF 5.0 / CFME 5.11.
what fields do you use in the report? Would it work for rhev
Created attachment 1703085 [details]
Report: VMs Requiring Restart
Created attachment 1703086 [details]
Example report output
This and the previous screenshots show a custom report and a report result
with VMS that need a restart.
Though this was done on current ManageIQ master, as I wrote before, the schema
changes are not present in current ivanchuk (CFME 5.11).