Confused with allocation pools

[PAckermann]

Hi,

we have a cluster runnig with 3 nodes of each 96GB RAM. We use this for vCloud exclusively. I get now some problems with the configured allocation-pools in vCloud-Director. We have 10 organisations configured, 2 allocation-pools, 2 resource-pools and the rest PAYG. For the allocation-pools I want to be able to overcommit, but exactly this is not working. In the particular case the ressource-settings for this OrgvDC are as follows:

20GB vRAM, 20% guaranteed

I thought, that would mean, that the customer get 4GB of vRAM reserved for him and can overcommit up to 20GB vRAM. The problem is now, the customer has 2 vApps running in there with a total of 4GB vRAM configured for the VMs. He tried now to upgrade one VM by 2GB vRAM, but gets the error-message, that this would exceed the vDCs ressources. I checked now all revervations and also current usage of vRAM in the cluster and there is plenty of RAM left. If I count all reservations of my allocation- and ressource-pools, I get a total of 83GB, the actual usage is 19GB. What am I missing here?

[PAckermann]

Solved it myself after some fiddling around. The problem was, that the customers storage-allocation was 100% in use. It seems, that vCloud-Director counts all configuration-data into this allocation and had no more space left to alter the configuration. This is in my opinion a bug that should be fixed as it is quite confusing and also not really clear to the user, that he shouldn't use all of his storage-allocation. In my case the customer had a 200GB-allocation and just provisioned 2 VMs with each 100GB HDD.