Created attachment 694577[details]
engine.log
Description of problem:
Quota exhaustion is remembered after first blocked action and thus not allowing any next action on the VM even resources of the VM are then below grade limit.
Running a VM with resources above grace (hard-limit) triggers quota resources exhaustion. This exhaustion is kept remembered. So after being blocked for first time, then when you redefine VM's resources (put them below hard-limit) and then you start the VM again, it is again blocked.
2013-Feb-07, 16:52
Cluster-Quota limit limit exceeded and operation was blocked. Utilization: mem:0%, Requested: mem:101.07% - Please select a different quota or contact your administrator to extend the quota.
and then...
2013-Feb-07, 16:52
Cluster-Quota limit limit exceeded and operation was blocked. Utilization: mem:101.07%, Requested: mem:50% - Please select a different quota or contact your administrator to extend the quota.
Version-Release number of selected component (if applicable):
sf6
How reproducible:
100%
Steps to Reproduce:
1. on a DC which is in 'audit' quota mode, assign to a user (AD) role 'VmCreator'
2. create some quota with following setup:
* memory 1024 (soft 60% => 615)
* cpu 3 (soft 60% => 2)
* disk 10G (soft 80% => 9)
3. assign to previously used user following role on quota - PowerUserRole
4. assign this user to the quota as consumer
5. login with the user into User Portal
6. create new VM, the VM should inherit valid quota with following resources
* memory 1035
* cpu 1
7. run the VM via Run Once with cdrom booting (just to put VM up)
8. check in Admin Portal quota usage
9. run the VM via Run Once with cdrom booting (just to put VM up) with memory 512, cpu 1 (this is below hard-limit)
Actual results:
VM is blocked even it is not overpassing quota hardlimit, it is blocked because previous quota exhaustion is remembered.
Expected results:
If nothing is using quota, then I should be able to run a VM with resources below hardlimit even previous blockage.
Additional info:
Created attachment 694577 [details] engine.log Description of problem: Quota exhaustion is remembered after first blocked action and thus not allowing any next action on the VM even resources of the VM are then below grade limit. Running a VM with resources above grace (hard-limit) triggers quota resources exhaustion. This exhaustion is kept remembered. So after being blocked for first time, then when you redefine VM's resources (put them below hard-limit) and then you start the VM again, it is again blocked. 2013-Feb-07, 16:52 Cluster-Quota limit limit exceeded and operation was blocked. Utilization: mem:0%, Requested: mem:101.07% - Please select a different quota or contact your administrator to extend the quota. and then... 2013-Feb-07, 16:52 Cluster-Quota limit limit exceeded and operation was blocked. Utilization: mem:101.07%, Requested: mem:50% - Please select a different quota or contact your administrator to extend the quota. Version-Release number of selected component (if applicable): sf6 How reproducible: 100% Steps to Reproduce: 1. on a DC which is in 'audit' quota mode, assign to a user (AD) role 'VmCreator' 2. create some quota with following setup: * memory 1024 (soft 60% => 615) * cpu 3 (soft 60% => 2) * disk 10G (soft 80% => 9) 3. assign to previously used user following role on quota - PowerUserRole 4. assign this user to the quota as consumer 5. login with the user into User Portal 6. create new VM, the VM should inherit valid quota with following resources * memory 1035 * cpu 1 7. run the VM via Run Once with cdrom booting (just to put VM up) 8. check in Admin Portal quota usage 9. run the VM via Run Once with cdrom booting (just to put VM up) with memory 512, cpu 1 (this is below hard-limit) Actual results: VM is blocked even it is not overpassing quota hardlimit, it is blocked because previous quota exhaustion is remembered. Expected results: If nothing is using quota, then I should be able to run a VM with resources below hardlimit even previous blockage. Additional info: