Zip Backup to CD - data backup software

Zip Backup to CD - Support Forum

Backup forum overview

Subject: Differencial Backup Please

By: Grant

16/05-2006 01:51

I have had this software since 2002 and still can not work how to do a Differencial backup. I have tried every thing I can think of.

All I want to do is have a base point say 1st may 2006 and then do a backup 8th May that includes all new and changed files at that point from 1st of 1st May. Then again on the 15th may still with reference point 1th May.

It uses a bit more space but I find it better if something goes wrong.

By: Bruce Mutton

31/05-2006 01:53

Grant
I have been using this software to do differential backups since about 2002. Works fine BUT the instructions are contradictory and misleading. Having said that look up differential in the help file. It explains how to do it in detail if you are using the catalog. If not, it is quite straight forward.
You need to set up your FULL backup job, save it, then save it as a DIFF backup job, which will be identical in all respects except that you back up CHANGED files, not ALL files, clear ABit will be UNchecked, and you probably want a different destination file name.

My only gripe is that to do differential backups using the catalog requires manual intervention (according to the instructions - not tested) at each backup cycle. Incrementals are fully automatic, but tooo many zip files if you need to restore without the benefit of the catalog.

By: Bruce Mutton

19/06-2006 11:42

Further...
Since I upgraded to version 3.19.8.504 about a month ago, to take advantage of the catalogue and DVD sized zip files, all my DIFFERENTIAL backups have turned out to be INCREMENTALS, which is becoming very tedious.

I have reviewed the on-line and as-shipped help files carefully, and run quite a number of tests, to no avail. Only FULL and INCREMENTAL backups seem to be produced.

Perhaps Datahjaelp can offer to review the settings I have used?

Please...

By: Datahjaelp

19/06-2006 11:47

Please, let me have your job file (jobname.dat) and I will look at it.

By: Grant

20/06-2006 04:56

Interseted I the result.

By: Datahjaelp

20/06-2006 13:37

The catalog is a file which contains information on all the files in the zip files which it was used together with or which was manually added to the catalog. The catalog can be used to determine which files needs to be backed up or to find the name of the zip file in which a file where archived. The catalog is enabled if the Cat field are checked, if the catalog name is left blank the catalog with the same name as the backup job will be used. The catalog is updated when you make a backup and the Cat field is checked. It is possible to specify more catalogs to be updated. Separating more catalog names in the catalog field with the "|" character will result in that the only the first catalog is used for checking if changed and newer than, and all the following catalogs are updated. The first catalog is not updated so it has to be listed twice if it have to be updated. If you would like to make a backup of the files which was changed or are new you have to select changed and check the Cat field and make sure that the selected catalog name is the one with the appropriate information's. It is possible to use the same catalog file for different backup jobs, the catalog file names to be used with a backup job is stored in the backup job file.

The full backup should have the following catalogs:
data_diff.cat
The data_diff.cat is updatet with informations from the full backup.

The differential backup should have the following catalogs:
data_diff.cat|data_inc.cat

The data_diff.cat is used during the backup if the changed and use cat is checked. It is not updated as there is a second filename.

The archive bit is working if you have the right to set or clear the archive bit. The problem is that in many cases you do not have the right to change the archive bit and in this case it will not work.

I will come back with a solution with running a job before the full job is started to delete the data_diff catalog.

By: Bruce Mutton

21/06-2006 00:14

Datahjaelp (Rene)
Thank you for your prompt reply.
Reading between the lines of your post you seem to be implying that I should not use a single catalog file for either backup cycles, or that if I do, I need to operate two catalog files in tandem to achieve a differential backup, and delete one of them at the change of each cycle. Please confirm this is what you intend after you get back on the 29 June.
In the meantime I will persist with the approach I have started out on, as I am making some progress.
I don’t think I have any file access permission issues or issues with archive bit setting and resetting, but I have not rigorously tested this yet.

For other readers background information, I have set up two daily backup cycles of one month’s duration as follows.
A FULL/DIFF cycle for dynamic data (‘Dyn’- compressed data size less than 4GB -1dvd) and
a FULL/DIFF cycle for static data (‘Static’- compressed data size about 15GB -4dvd’s)
This requires four separate job files (of the form jobname.dat) that Datahjaelp have reviewed for me, and responded with the above post.

I have opted to go for the approach implied in the last paragraph of Datahjaelp’s post (and also in the help file) “It is possible to use the same catalog file for different backup jobs…” because of the benefit I perceive in easier setup of the job file and only ever having to look in one catalog to find file(s) to be restored. On the face of it running multiple catalogs appears to me to make it harder to locate particular files for restore at some future time. Over the past month I have used this setup quite a lot to find and restore most recent versions and old versions of files from both backup cycles and it appears to work perfectly, even if the user interface is not the easiest to work with. The problem I am having is that of my DIFF cycles turning out to be INCR cycles instead.

Now I think I have a breakthrough!

My earlier post was slightly in error, as I find that on closer inspection it seems that only one of my backup cycles is misbehaving and producing INCREMENTAL data sets. The ‘Dyn’ set seems to be behaving and producing DIFFERENTIAL data sets as intended, at least for the sub set of files that I have been able to track.
On printing out a screen dump of the application with each of the job files loaded, I find that what I would have thought was a mistake in my DynDIFF.dat job seems to be the setting that allows it to work as intended. (The ALL and CHANGED settings are reversed)

So, to summarise settings that seem to enable use of the catalog with multiple or with a single set of backup cycles…
My hope is that the ‘After’ date rolls over automatically with each monthly backup, and although this seemed to work once, I don’t see how the application can track this properly between multiple job files. I wait in hope for the next moth to roll by to see what happens…
-JOB FILE-
Setting: FULL DIFF

Use Last Runtime As Newer Than: ON OFF
All: ON ON
Changed: OFF OFF
After: OFF ON
Clear Abit: ON OFF
Cat: ON ON
Cat File: Mybackup.cat (in both cases)

(I know the archive bits are not used by the application, but it is helpful to clear them so that they accurately reflect the backup status of the files)

Please let me know if there is a problem with the way I am trying to use the software.

By: Bruce Mutton

21/06-2006 00:18

Drat
The table near the end of my post got scrambled.

It was meant to be a three column table, the first column is the application setting, the second the setting for the FULL.dat job file, and the third the setting for the DIFF.dat job file.

By: Bruce Mutton

14/07-2006 11:20

In my post above I mused;
"My hope is that the ‘After’ date rolls over automatically with each monthly backup, and although this seemed to work once, I don’t see how the application can track this properly between multiple job files."

I was correct, the application does not track roll-over dates across job files, so my while my approach works with manual user intervention each cycle roll over date, it can not work automatically.

Am working on the approach that datahjaelp seems to be suggesting using two catalog files, and with some experimentation it looks as though it might work, once I weed out the ambiguities (errors?) in the above post and the instructions.
Having a second sacrificial catalog file that is deleted regularly does seem to be self defeating. A large chunk of ones backup history is intentionally removed.
Bizzare!
I guess I may have to accept the lessor evil of a FULL/INCREMENTAL backup cycle...

By: Bruce Mutton

14/07-2006 23:39

... or more sensibly a FULL/DiFFERENTIAL backup cycle without the benefit of catalog's as I had been successfully, happily and reliably doing for many years before the tempting features in version 3.19.8.504.

After all, windows XP provides an adequate directory tree based interface with which to manually browse zip files that in a number of ways is superior to the text list interface in Zip Backup to CD...

By: Grant

15/07-2006 00:35

Bruce

I just gave up and bought Zipbackup which easily does what I want and more without the frustration. The the backup I want should not have to be rocket science like it is with this program. Pity after 4 years.

Grant

By: Bruce Mutton

16/07-2006 10:35

I agree Grant. I trialled Zipbackup for the trial period a couple of months ago, it worked flawlessly, has an idiot proof mode (although very basic functionality in that mode), the instructions were fairly comprehensive and were accurate, the backup logs are much easier to read, and the advanced mode worked like a dream.
Probably more 'usability' advantages as well.

Still, I thought ZipbackuptoCD was a bit more comprehensive, in terms of systemstate backups and some other things that escape me now.

Again. You are right, this is much much harder than it should be. You could do worse though. Geniebackup is highly rated, but is even harder to beat into shape than Zipbackup to CD.