Lets pretend, shall we?

1. Users are actively using the system, entering data and what-not.

2. Level 0 ontape is started at 4:30pm.

3. All users are logged off at 5:30pm.

4. Ontape completes successfully at 5:45pm.


Question:

Was all of the data backed up? Or was some data potentially missed
since users were entering data for that 45 minute period that ontape
was running?

I say the latter, but now I'm hearing that ontape would have picked up
everything.

Anyone know the correct answer?

Thanks!

Yep, the backup is consistent to the first check point that is done
after the ontape is issued. ontape forces a checkpoint for this
purpose.

If it tried to keep up with the changes that happened while it was
running you might get into a situation where ontape could never finish
because updates were happening faster than ontape could record them.
>From what I understand the before images of pages that are changed
while ontape is running and that it hasn't backed up yet, are put into
the tempdb spaces so that ontape can back them up in their original
state without affecting update peformance much. Of course if you make a
lot of changes you can have temp space issues.

natebsi@gmail.com wrote:
> Lets pretend, shall we?
>
> 1. Users are actively using the system, entering data and what-not.
>
> 2. Level 0 ontape is started at 4:30pm.
>
> 3. All users are logged off at 5:30pm.
>
> 4. Ontape completes successfully at 5:45pm.
>
>
> Question:
>
> Was all of the data backed up? Or was some data potentially missed
> since users were entering data for that 45 minute period that ontape
> was running?
>
> I say the latter, but now I'm hearing that ontape would have picked up
> everything.
>
> Anyone know the correct answer?
>
> Thanks!