I know that the block sizes for backup a backup job are variable and linked to the settings in the backup job (local/LAN/WAN). What I didn't know is that this could be variable within the same job on restore. In this example I see a restore happening of the first disk of a machine at 100k block size on average but on the second drive we see block sizes of +220k off course with different IOps/MBps results on the array. Can someone help me understand this behaviour? I would be good to know if my array for example handles 100k better that I could tweak that to get the best/fastest results on restore.
Hi, Hans. This is due to different compression ratio for data stored in the first and second disk. Actual block size is static within the job, this is the block size we use to process raw data from source virtual disk during the backup. However, what this data compresses to, will be different from block to block. Thanks!
