BE 2012 True Image Backup (Using Catalogs) takes long to start

[SFietzek]

Hi,

I just upgraded to BE 2012. Now I want to use the True Image Restore feature for the first time. The appropriate license is installed.

So I created a backup job (full + incremental) for one volume (600 GB, > 1 Mio files) and selected the "use catalog" method which is required for the True Image Restore feature. When the full backup job starts it takes a lot of time until I can see a backup progress. After the job started and the VSS snapshot is created it stay at 0 Byte and 0,00 MB/min for moren than an hour until it really starts backing up files.

I check the Remote Agent Utility at the host and can see that something is working. It seems that all files are enumerated before they are backed up.
I also checked that the NTFS journal is active on the volume I want to back up.

The behaviour seems to be the same on all my hosts and volumes. The bigger the volume is the longer it takes until the backup really starts.

Why does this happen? What am I doing wrong? Any help would be appreciated.

[pkh]

It seems that all files are enumerated before they are backed up.

This is correct because True Image backups needs to capture the state of the folders.  For example, when you do a True Image restore, deleted files will not be restored, unlike normal restores.

How much longer are these backups compared to backups which are not True Image backups?

[SFietzek]

I understand that Backup Exec needs to capture the state of the folders but why does this have to be a separate step?

At first I take full backup. This means everything has to be backed up. Backup Exec sees all folders and files anyway. --> No separated step to capture the state needed.
The there will be diffential or incremental backups. Backup Exec should use the NTFS journal. As far as I understand every file system change will be in the journal. So processing the journal leads to all relevant information. --> No separated step to capture the state needed.

The only time a separate step would make sense is if there was no NTFS journal and Backup Exec has to compare the catalogs against the file system. But I hope in my case the journals are used. Can this be validated in any way?

Aditionally it would be fine to mention this step anywhere in the documentation and show it as the current state of the backup job. This would be much clearer than showing "Backup..." and not showing any progress for a large amount of time.

To answer your question:
I backed up a volume with 682 GB and 3.9 mio. files to disk. The whole job took 8 hours. About 1 hour or something more there was no progress so this must be the initial state capturing you mentioned.