change log details for "changes only"

[userX]

The update history log shows that for v 2.78

1. Added the ability to create incremental backups in addition to differential.

Where can I find the information that explains at what point in the back-up procedures, using IFW or IFL, a user would need to make the decision as to which of these two types of backup to make. When I look at the first IFW dialogue with version 2.79, I see options for Backup (Full) and for Backup (Changes Only), but it is not clear which type I would be using if I chose the latter option. Also, when I look at the settings dialogue after clicking the Settings button, I don't see it in the options anywhere.

2. The image file created by this version requires an Image version 2.78 or later to restore.

Can someone clarify this statement for me?

3. Added /COMBINE command-line-only operation to combine changes-only backups to single full backups.

Do I understand correctly that this command would allow me to consolidate several sets of "changes only", along with the base to which they correspond, into one single complete backup at that point in time, and that I could then use this as a new base going forward for making some more "changes only" backup sets?

[TeraByte Support(PP)]

1. The "point" is where you select the base file for the new backup. If you select an existing differential or incremental backup as the base an incremental is created. For example, if you want to create an incremental chain, create a full image, then a changes only backup based on the full, then another changes only backup based on the previous changes only backup, and so on.

Full - Base image
IncA - Based on Full
IncB - Based on IncA
IncC - Based on IncB
etc.

2. You would need to use 2.78 or later to restore any backups created with 2.78 (unless you used the Backwards Compatible option). So, normally, you wouldn't be able to restore an image created with 2.78 using 2.77 or earlier. It's recommended to use the current version of IFW (currently 2.80).

3. Correct. You can combine the chain (from a specified incremental back to the base) into a new full file.

[DrTeeth]

How can one tell which files (diff or incr) are which from the
filenames?
--

Cheers

DrT
______________________________
We may not be able to prevent the stormy times in
our lives; but we can always choose whether or not
to dance in the puddles (Jewish proverb).

[userX]

So when I am making "changes only" backups and I point at the original "whole disk" backup data as the base, my new backup data would be made as a differential backup data file, and any previous differential backup data file created from this same base data could now be deleted because restoring from differential would require only the original base data and the most recent differential data. Is that right?

In contrast, creating a new "changes only" backup data file by using a base of not the original "whole disk" data but of either type of "changes only" backup data files created from it now would create an incremental backup data file; and in using incremental backups I in order to restore my data, I would need not only the original "whole disk" backup data file but also ALL of the "changes only" data files that were used for bases during previous backups. Is that right?

[TeraByte Support(PP)]

DrTeeth wrote:
> How can one tell which files (diff or incr) are which from the
> filenames?

I usually just indicate in the filename which it is.

[TeraByte Support(PP)]

userX wrote:
> So when I am making "changes only" backups and I point at the
> original "whole disk" backup data as the base, my new backup data
> would be made as a differential backup data file, and any previous
> differential backup data file created from this same base data could now be
> deleted because restoring from differential would require only the original
> base data and the most recent differential data. Is that right?

Yes, but the base image only needs to be a full image -- it doesn't need to be of the whole disk. For example, you could make a full image of a data partition and then create differentials/incrementals based on that full.


> In contrast, creating a new "changes only" backup data file by
> using a base of not the original "whole disk" data but of either
> type of "changes only" backup data files created from it now
> would create an incremental backup data file; and in using incremental
> backups I in order to restore my data, I would need not only the original
> "whole disk" backup data file but also ALL of the "changes
> only" data files that were used for bases during previous backups. Is
> that right?

Yes (same thing as above applies for the base image).

Restoring a differential requires the base image (full) and the desired differential image. Restoring an incremental requires the base image (full), the incremental being restored, and all incremental images in the chain between the base and the image being restored.

[userX]

Thank you, Paul. This outstanding program just keeps getting better and better. I sincerely think that you should market your quality control and development process itself, and maybe some other vendors could learn a few things about how to create customer loyalty and a standard for excellence.

[DrTeeth]

On Tue, 5 Mar 2013 12:26:05 PST, just as I was about to take a herb,
TeraByte Support(PP) disturbed my reverie and wrote:

>I usually just indicate in the filename which it is.

Call me old fashioned, but as I always use the filenames that IfW
provides, I would like IfW to do that automatically. In fact, I expect
it from a program with the quality of IfW .
--

Cheers

DrT
______________________________
We may not be able to prevent the stormy times in
our lives; but we can always choose whether or not
to dance in the puddles (Jewish proverb).

[userX]

"...I always use the filenames that IfW provides"

You can continue doing that if you really need to for some reason, but, honestly, it sounds like you are mixing up the apples with the oranges, and that is no way to keep track of something as important as your backup files. They ought to be top priority for your attention to care considering what they represent. Just put your increments in one folder and your differentials in another if you have to do so to keep them straight. The date is right there on them. Why are you mixing them anyway? It's better to find a procedure that is efficient and follow it consistently. At least that is how I try to work it.

[userX]

One thing that occurred to me about what you asked is that there may be a workaround if you should have made both a differential and an incremental. It might be possible to combine them, based on my first question and point number three:

"Added /COMBINE command-line-only operation to combine changes-only backups to single full backups"

I would think that, if you prefer to have both types, you could later use this options to put them together. Then consolidate the whole set of changes with the original base. Would that not work?