Jump to content

Why is the data backed up again?


Recommended Posts

Server is HP proliant with Windows 2003 server, Retrospect 7.6.123, driver update 7.6.2.101.

 

Disk backup set.

 

The client has four volumes:

XP, which is a bootable XP volume with client 7.6.107 installed.

Vista, which is a bootable Vista volume with client 7.6.107 installed.

Data1, which has data files on it.

Data2, which has data files on it.

 

The client is added to the server twice under different names, once when booted XP and once when booted Vista.

 

Scenario:

Client is XP booted. Backup the client using Recycle backup.

Boot client on Vista. backup the client using Normal backup.

What happens is that all data on all volumes is backed up again.

While I understand that the Registry is completely different and gets backed up twice, NONE of the files on Data1 and Data2 has changed.

Yet is every file backed up again. Why?

Does Retrospect see the metadata as changed?

But the info on the volumes hasn't changed.

 

No, it's no weird setting. Normal backups work just fine on all other clients and on this client: The next backup backs up the changed/new files as expected. (But they are backed up twice: Once booted on XP and once for Vista.)

Link to comment
Share on other sites

What do you mean by meta data?

 

The meta data in NTFS? Should Retrospect compare the entire file, like doing md5 or something?

 

Thanks

Oliver

 

Vista may be making the files appear different to Retrospect, specifically the meta data may be different.

 

Retrospect looks at the name, size, date/time, meta data

 

Link to comment
Share on other sites

Should Retrospect compare the entire file, like doing md5 or something?

 

Retrospect does not do block-level deduplication. !!!

 

In addition, OS files do get recopied.

 

According to the manual, it relies on the name, size, creation date and modification date. There is no reference to volume identifying data. I don't know where the meta-data, whatever that is precisely, fits in.

 

There are also a couple of options governing the degree to which it checks, namely, "Match source volume to Catalog File", "Don't add duplicates to Backup Set", and "Match only files in the same location".

Link to comment
Share on other sites

There are also a couple of options governing the degree to which it checks, namely, "Match source volume to Catalog File", "Don't add duplicates to Backup Set", and "Match only files in the same location".

"Match source volume to Catalog File". Checked.

"Don't add duplicates to Backup Set". Checked.

"Match only files in the same location" NOT checked.

(These are the default settings.)

Link to comment
Share on other sites

This is the exact type of problem I am having. I had a data drive go bad - replaced it - was able to move the files from the bad drive to the new drive - and retrospect insisted on backing them all up again. I have the same check boxes checked as Lennart which should stop this, but it doesn't. So there is something Retrospect is seeing causing it to reback up all these files when in reality none of them have changed. The disk info should be totally irrelvant, it should be file info only that is considered.

 

So what is Retospect seeing that is triggering this? Whatever it is, it's wrong.

Link to comment
Share on other sites

maurice - what exactly is block level de-duplication?y

Wikipedia is your friend:

http://en.wikipedia.org/wiki/Data_deduplication

http://en.wikipedia.org/wiki/Block_(data_storage)

 

In short, it means a backup program backups the changes on a lower level than the file level (in this case block-level), instead of backing up the whole file.

 

Say you have a 25MB large TIFF file. You change one pixel. It will only backup the tiny couple of bytes that actually changed and will be able to restore later from the initial backup plus the recorded changes on the block level. This saves a lot of space & time.

 

Currently Retrospect can only backup the whole file, even if just a pixel is changed. So taking up another 25MB in your storage as opposed to just the difference on the block-level (mere bytes in this case).

 

Block level data deduplication is the new revolution in backup technology.

Link to comment
Share on other sites

Re: Wikipedia and Google.

 

Right. Thanks. I use them of course. But I really want to hear from Retrospect on it because I have no idea what they consider or look at in their backup process related to metadata or how it's treated.

 

Somehow Retrospect is considering something a "new" file when it really shouldn't.

 

--Randy

Link to comment
Share on other sites

But I really want to hear from Retrospect on it because I have no idea what they consider or look at in their backup process related to metadata or how it's treated.

 

Somehow Retrospect is considering something a "new" file when it really shouldn't.

It's all configurable in the backup preferences, and is explained in the manual.

 

Personally, I wish that the interface was better for choosing what parameters were used for matching files to be backed up. Perhaps a checkbox for metadata issues to be ignored, such as: Time difference less than 1 second (matters if file metadata is being compared between different underlying filesystems, such as NTFS and Linux, etc.), permissions, owner, group, etc.

 

Russ

Link to comment
Share on other sites

Russ,

 

Believe me, I'm familiar with every backup pref and NONE of them solve the problem.

 

As you said in your second paragraph, it's way too sensitive in some areas somehow - who knows where - and those areas need to be brought out front into GUI options.

 

I'm windows, so with all their MAC issues now I don't expect this to get addressed for awhile unfortunately. It just MUST be addressed in the next windows release, IMHO.

Link to comment
Share on other sites

I think the problem is in the client.

 

I just had to move a Retrospect installation in another client, Windows XP backing up in an external USB HD.

 

I backup up the same data via SMB and local and it worked as expected.

 

Will test backup up a client with the unchanged data on that client to confirm it.

 

What I mean is:

- the same data backed up using another client, de-duplication does not work

- the same data backup up using SMB or local, de-duplication does work.

 

 

HTH

Oliver

 

 

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...