I have questions about Windows 7 Backup & Restore. From everything I am reading, it would appear that the program performs as advertised, but some of the traits I am witnessing just don't make any sense to me, and when it comes to Backup/Restore (B/R), things need to make sense.
Q1: I have a machine that is configured with three (3) accounts: one is a privileged account (admin), and the other two are normal non-privileged accounts (acct1 and acct2 for reference purposes). I perform the backup from the admin account, backing up the entire machine (all accounts). The backup completes without error. If I then Restore Files to view what was saved, I am shown the admin account, and acct1 as available to restore. Acct2 is not visible! Why not? Am I not allowed to restore this account as admin? If it is a permissions issue, then why can I see and select acct1? If I logout of admin and login to acct2, and run B/R from there, the admin and acct2 backups are visible, but not acct1. So I know the data got backed up, but why am I not able to see all the backup accounts when running as admin? If I had a machine with 15 accounts, does this mean I would have to login to each account to restore its files when rebuilding after losing a drive?
Q2: Backups are tagged by date. If I run B/R daily, it appears that the backup is retagged with a date range. If I wait a number of days to perform the second backup, I may get a second backup entry with the new date. What algorithm is being used to determine when to extend a backup, and when to create a new entry? Is there a trigger that I can use that will force the creation of new entry, and not extend an existing entry, or vice-versa?
Q3: Everything suggests that with multiple backups, the oldest backup represents a full set of files, and subsequent backups are incremental. Exactly what is meant by incremental. I would expect this to mean that any file that has not changed since the last time it was backed up would not be part of a new backup. Or stated the other way, any file that has changed since the last time it was backed up would be part of the new incremental backup. If I have 4 backups, will B/R examine all 4 to determine whether a file needs to be included in the new incremental backup (I assume so)? If I have 15 backups will B/R examine all 15 to calculate the new incremental (I assume so)? If I decide to manage space and delete the oldest backup, will a subsequent backup operation resave all my old files that have not changed since the first backup but are now missing from the backup set (I assume so)? If so, then any space gained by deleting the original backup will disappear after the new incremental completes. Would it be better to delete only the older incremental backups to free up space?
Q4: If I were writing this program, I would expect that contained in each backup is an index to all the saved files along with their dates of existence. I would have to scan ALL backups to create an index of ALL files with their respective save dates. I could then compare the current file path and date with the name and date in the archive to determine whether the current file needed to be included in the new incremental. If this is the case, is there some point where deleting some of the older backups will improve the overall efficiency of the backup operation by reducing the number of backups that need to be scanned to build the current backup file list? If so, what is a good rule of thumb for the number of backups to retain to preserve program efficiency?