Created attachment 937861[details]
screenshot
Description of problem:
If vm stopped in state 'Waiting for Launch', engine not update host 'Max free Memory for scheduling new VMs' in webadmin
Version-Release number of selected component (if applicable):
rhevm-3.5.0-0.12.beta.el6ev.noarch
How reproducible:
Always
Steps to Reproduce:
1. Create vm and run it
2. Stop vm, in 'Waiting for Launch' state
3.
Actual results:
engine not update host 'Max free Memory for scheduling new VMs' in webadmin
Expected results:
engine update host 'Max free Memory for scheduling new VMs' in webadmin
Additional info:
I also attach screenshot and database line:
engine=# select mem_available from vds where host_name='master-vds10.qa.lab.tlv.redhat.com';
mem_available
---------------
22984
(1 row)
Also REST not updated and have <max_scheduling_memory>3807379456</max_scheduling_memory>
(In reply to Artyom from comment #2)
> yes, if you need I can try to check other cases, when it not first vm.
If I understand the problem correctly, the problem is that in this particular case we don't decrease the pending memory, but the committed memory.
So if there was no other Vm running on the host, the 'max_scheduling_memory' would not be decreased (because there was no 'committed memory' before).
If there was another VM running on the host its committed memory will be decreased - the bug will still exist but you'll see that the max_scheduling_memory is being decreased (I expect you'll see that the pending memory stays the same and the committed memory is reduced).
It will be great if you can verify it.
Created attachment 937861 [details] screenshot Description of problem: If vm stopped in state 'Waiting for Launch', engine not update host 'Max free Memory for scheduling new VMs' in webadmin Version-Release number of selected component (if applicable): rhevm-3.5.0-0.12.beta.el6ev.noarch How reproducible: Always Steps to Reproduce: 1. Create vm and run it 2. Stop vm, in 'Waiting for Launch' state 3. Actual results: engine not update host 'Max free Memory for scheduling new VMs' in webadmin Expected results: engine update host 'Max free Memory for scheduling new VMs' in webadmin Additional info: I also attach screenshot and database line: engine=# select mem_available from vds where host_name='master-vds10.qa.lab.tlv.redhat.com'; mem_available --------------- 22984 (1 row) Also REST not updated and have <max_scheduling_memory>3807379456</max_scheduling_memory>