VM capacity and prism capacity are different | Nutanix Community
Skip to main content
Hi,

I got a question during the test.



Window VMs(C: drive) , linux VMs(df -h) capacity different those capacity in the prism(VM-table).



Not all VMs are like this, only a few are like this.



These VMs erased a lot of data and got a lot of capacity.



The Curator has since been executed, but the capacity has not decreased inf prism-vm-table.



Why can't Prism return a VM's erased data capacity value?



How do I get Prism to read this capacity?



Window -> Server 2016

Linux -> Centos 7.X



Thank you.

@Hong 

The reason you don’t see an increase in free storage space has to do with the logic of data removal at different layers of the environment.

When you remove data at the Guest OS layer it is not overwritten with zeros but instead is marked as available for overwriting (the terms can be argued based on the guest OS type but the general principle remains roughly the same – the data remains). This means that while guest OS file system may show ample free space this is not necessarily reflected on the hypervisor or storage layers.

Assuming the underlying hypervisor is AHV, the solution would be to zero free space. As AHV does not write zeros to the storage, zeroed space would be freed up when discovered by Curator.

You could leverage sdelete on Windows.

https://docs.microsoft.com/en-us/sysinternals/downloads/sdelete

For Linux I would suggest to read this https://unix.stackexchange.com/questions/44234/clear-unused-space-with-zeros-ext3-ext4

 

Please let me know if the above makes sense or if you still have any questions.