Differential backup duration

[The Seeker]

For me, performing a differential backup seems to take as long as a full backup. I was always under the impression that a differential backup should take less time than a full. This being the case, there seems to be no point at all for me to perform one.

For example, last night I performed a full backup. This morning, I deleted a few large files then instigated a differential backup, which was set to take just as long as last night's full one. My full backup is contained on an external HDD, and I planned to store the differential on the same drive. Am I missing something here?

[TeraByte Support(PP)]

The main benefit is that differentials are usually a lot smaller than full images. Keep in mind that the entire source is checked for changes. This means that the same data read to create the full image is read to create the differential.

[TAC109]

On Tue, 6 Mar 2012 04:00:39 PST, The Seeker wrote:

>For me, performing a differential backup seems to take as long as a full backup. I was always under the impression that a differential backup should take less time than a full. This being the case, there seems to be no point at all for me to perform one.
>
>For example, last night I performed a full backup. This morning, I deleted a few large files then instigated a differential backup, which was set to take just as long as last night's full one. My full backup is contained on an external HDD, and I planned to store the differential on the same drive. Am I missing something here?
>
If you use the option to validate the backup afterwards (not
byte-for-byte), the validation phase will be much quicker on the
smaller diff file.

[The Seeker]

Keep in mind that the entire source is checked for changes. This means that the same data read to create the full image is read to create the differential.

Ah, that makes sense. Thanks for replying