Describe the solution you'd like
When using snapshotReserve, the following parameters should be set ( or options to set them should exist ) :
auto-delete: true
trigger: snap_reserve
The behaviour right now is not very user friendly (k8s users friendly). K8s users dont understand that they dont have space anymore in their volume because of snapshot overspilling.
When using snapshotReserve to 50% for exemple, with the two parameters above, user are garantied 10 GB of space if they asked 10 GB. Snapshot will not spill in the user's space.
Describe alternatives you've considered
We created a script to fix those parameters on volume on which it is incorrect. We shall run this script daily to make sure new volumes created by trident are standard.
Additional context
Here is a volume created by trident and the snapshot autodelete parameters :
SuperNAS::> snapshot autodelete show -volume dev_vert_pvc_6004add8_e8a2_42db_8e63_8ee4c5e68876
Vserver Volume Option Name Option Value
--------- -------------- ------------------ ---------------------
SuperVServer dev_vert_pvc_6004add8_e8a2_42db_8e63_8ee4c5e68876
Enabled false
Commitment try
Trigger volume
Target Free Space 20%
Delete Order oldest_first
Defer Delete user_created
Defer Delete Prefix (not specified)
Destroy List none
SuperNAS::> volume show -vserver SuperVServer -fields percent-snapshot-space -volume
-volume -volume-style -volume-style-extended
SuperNAS::> volume show -vserver *Dev* -fields percent-snapshot-space -volume dev_vert_pvc_6004add8_e8a2_42db_8e63_8ee4c5e68876
vserver volume percent-snapshot-space
---------------------- ------------------------------------------------- ----------------------
SuperVServer dev_vert_pvc_6004add8_e8a2_42db_8e63_8ee4c5e68876 50%
SuperNAS::>
Describe the solution you'd like When using snapshotReserve, the following parameters should be set ( or options to set them should exist ) :
The behaviour right now is not very user friendly (k8s users friendly). K8s users dont understand that they dont have space anymore in their volume because of snapshot overspilling.
When using snapshotReserve to 50% for exemple, with the two parameters above, user are garantied 10 GB of space if they asked 10 GB. Snapshot will not spill in the user's space.
Describe alternatives you've considered We created a script to fix those parameters on volume on which it is incorrect. We shall run this script daily to make sure new volumes created by trident are standard.
Additional context
Here is a volume created by trident and the snapshot autodelete parameters :