This content has been marked as final. Show 5 replies
961469 wrote:Some confusion on above scenario, As i understood correctly the stright forward question is, Can you create manual restore points ? If so Yes, but create a restore point with Guarantee , You can refer below link for more details
I really dont think this is possible so expecting a no but asking anyway as I cant find a definitive answer
Dataguard standby in physical standby mode
Open in standby snapshot mode for Read Write testing, this automatically create a snapshot point
Re-Open in physical standby mode which uses above snapshot point, querying the snapshots after open and we dont see it anymore.
My question is it possible to have a manual snapshot point created on the standby database which is persistent across snapshot and physical opens. So
Open standby in snapshot mode (auto snapshot point A)
do some work
take a manual snapshot (point B)
open in physical mode (point A, no longer available)
some time later open database in snapshot mode using Point B.
I cant test this til later in the week but Im guessing Point B will be just gone once physical is re-opened.
maybe I can clarify a bit better
1. Database in dataguard physical standby mode
2. Open in snapshot RW mode: Point A
3. Do work
4. Create restore point B
5. Open in physical mode which restores to Point A and synchs up with primary.
6. Some hours later at Point C I want to put database back into snapshot mode but want to start at point B, not at Point C which will be created when I use the dataguard broker to open
7. Open in physical mode at Point C (puts standby back in synch with Primary)
Im dubious as to whether point B will remain persistent after step 5 above.
I don't see how you could do that. In order to get back to point A which is the physical mode, you have removed point B. Then, data is applied from the primary, you stop it and want to get back to point B. That means that you'd have to go back to point A first, then up to point B?
Maybe i'm missing something or not understanding exactly what you are saying but i'm not sure it's possible. Would be interested to know if it is.
We have physical standby DBs which we open for read write use every week. We create a guaranteed restore point, open the DB for use, then flash it back once we're done. Then start redo apply again. We obviously lose any work done while the DB is up, even if we did create a second guaranteed restore point it would be gone because we flashed back to a point in time before that 2nd restore point.