How does snapshot is created on the disk

[fahmi_2]

I understand theoricaly what is snapshot and how we can use it to backup and restore data.
But i would like how does technology is implemented.

Is there anyone how can explain it to me.

[Uwe Zessin]

Hello fahmi,

imagine what a server sees: a list of blocks of data. The storage array maps these blocks to the physical disk drives.

If you request a snapshot, the storage array creates a second, identical mapping for the new view that is presented to a server.

So you now have identical view through the original and the snapshot volume to the same disk blocks.

For your backup, you don't want that the snapshot view changes when your production server is updateing any data block. In that case the storage arrays holds off the write of the new data to the disks for a moment. It reads the unchanged data from the disks, looks for a free portion of space withing the array and puts the data there.

Next it updates the mapping for the snapshot view, so that your backup still has access to the old, unchanged data. Now it can write the new data from the server to the disks. This is called a copy-out operation.

Some storage arrays may do it a bit different by looking after free space for the production volume and updateing its mapping, but the effect is the same - the snapshot view does not change.