The nature of the data is very important for deciding whether to enable deduplication. Databases or media files, such as JPEG, MP3, and MOV, might not be the best candidates for deduplication, as they have very little or no duplicate data. Virtual machine boot image files (
vmdk files), user home directories, and file system with multiple copies of files are good candidates for deduplication. While smaller deduplication chunk size normally results into higher storage saving, it takes longer to deduplicate and requires a larger deduplication database.