The error is clear, there is no space on the san for snapshot.
The space used for snapshot is allocated on snapshot creation during the backup, as a % of volume to be snapp-ed so i thing 10 to 20% of the 40GB ( do not remember if on lefthand the snapshot are thin provisioned so maybe it is not the case ).
But in every case tre must be unallocated space on san.
you can set up the volume as thin provisioned ( at any time ) so that only the space used ( even once so delete do no reclaim the unused blocks ) wil be allocated.
Even this i do not feel confident using a single 40TB volumes for all VM, i often use volumes between 1 and 2 TB, shared between multiple VM ( i do nor find an hand limit on this ) and this has several advantages:
- corruption of 1 csv don not impact all vm
- lack on free space on one csv do impact all vm
- the snapshot are tiniest.
- less backup running concurrentry on the same csa ( better )
using thin provisioning give you the option to create other volumes and move the vm one by one.
I'm positive that this will solve your backup problem.
PS:i allocate a given percentage of space as a non thin provisioned volume NOT used to have readiliy available usable space in case the abuse of thin provisioning cause an overallocation of space.
Best regards.
- - - Updated - - -
The error is clear, there is no space on the san for snapshot.
The space used for snapshot is allocated on snapshot creation during the backup, as a % of volume to be snapp-ed so i thing 10 to 20% of the 40GB ( do not remember if on lefthand the snapshot are thin provisioned so maybe it is not the case ).
But in every case tre must be unallocated space on san.
you can set up the volume as thin provisioned ( at any time ) so that only the space used ( even once so delete do no reclaim the unused blocks ) wil be allocated.
Even this i do not feel confident using a single 40TB volumes for all VM, i often use volumes between 1 and 2 TB, shared between multiple VM ( i do nor find an hand limit on this ) and this has several advantages:
- corruption of 1 csv don not impact all vm
- lack on free space on one csv do impact all vm
- the snapshot are tiniest.
- less backup running concurrentry on the same csa ( better )
using thin provisioning give you the option to create other volumes and move the vm one by one.
I'm positive that this will solve your backup problem.
PS:i allocate a given percentage of space as a non thin provisioned volume NOT used to have readiliy available usable space in case the abuse of thin provisioning cause an overallocation of space.
Best regards.
