Network and Storage Protocols
Network and Storage Protocols
I have virtual machine (Ubuntu) on vmware. Datastore is connected by nfs. Data grow in the virtual machine and are cleared. But upon vmware shows that the virtual machine takes the maximum place. Everywhere thin provision. How to release a place? I know that it is possible to zeroing the empty place in virtually machine and to release with the esxi. Whether it is possible to automate it? Maybe the driver is?
Datastores created over NFS should automatically give you the freespace as data is deleted. Where do you see the space being occupied? At ESXi side of at NetApp side? Do you have a sample output of "df" command from DataONTAP CLI for both before and after deleting data? For VMFS datastores on NetApp LUNs, you may have to use the space reclamation feature of the Virtual Storage Console (VSC).
netapp:
vmware capacity of datastore:
vmware vm resources:
os #df -h
where is ~300Gb ???
You have a VMDK on an NFS datastore, which expanded to hold files which have since been deleted. You are looking for a way to reclaim that space, correct?
VSC has a space reclamation utility for NFS datastores, but I think it requires an NTFS file system in the vmdk.
With 8.3 or later inline zero elimination should give the space back to the volume if you zero fill the whitespace on that disk, but the vmdk itself will still appear to be full size.
Otherwise zero fill the whitespace, then storage vmotion it to another datastore with "thin provisioned" selected as the disk format.
Yes, I think I must zeroing space in Ubuntu OS, then reclaim it in ESXi. Whether simply I wanted to learn there is easier way...