I am wondering what happens in a multple SVM scenario when SVM A destroy a volume X. Then if SVM B creates a new volumes Y where blocks previously belonging to volume X are now belonging to Y. Is the data from volume X available through volume Y ?
Basically, how can I reassure a very cautious tenant about the security of his data. Is there official documentastion on this ?
Thank you for answering. I had the info on what you answered, but my question is left unanswered.
What happens when the second SVM reads a block that previously belonged to the first SVM ? Let's say its block 32 to the server using the LUN. What will block 32 contain when the server asks to read that block ? Will it contain all zeroes or will it contain data placed there previously when the block was part of a volume of the first SVM ? And how about if the new volume is a cifs volume ?
Here is my scenario. I already have a tenant using the equipment. I need, to introduce a second tenant to that same equipment because this is where I have the capacity. None of the volumes of the first tenant are encrypted as previously the equipment was dedicated. The first tenant has concerns on the privacy of his data if I introduce the second tenant.
Remember that the volume presented to the client/s is virtualised by ONTAP's WAFL layer to actual blocks on disk. The system will return exactly what clients have written to it - if they haven't written anything, it will return blank blocks.
Your client's data will remain on disk until the block is reclaimed and rewritten for another volume. There are ways to recover it in whole or part until this occurs. If they are very concerned about security, you could consider an upgrade to ONTAP 9.1 and enable NetApp Volume Encryption (NVE) for their volumes (if your controller supports it). This will introduce software encryption for their volumes.