Ingres Version: All OS: All Q: What purpose do the dump files serve when I checkpoint my database?
When running an online checkpoint, users are allowed to make changes to the database. This means that some of the table files (.t0x) may be modified during the course of the checkpoint. The checkpoint works by using the OS to back up the files. This means that if a table is modified, it’s corresponding table file MIGHT have been updated to reflect this change, then again it also MIGHT NOT have been updated; it all depends on if that particular transaction took place BEFORE or AFTER it’s table file was backed up. To make things more consistent, Ingres needs to be able to get a static ‘picture’ of the database at the precise moment that the checkpoint is run. To do this, it needs to know what transactions were run during the checkpoint; at recovery time (rollforwarddb), these transactions are rolled back, guaranteeing that the recovery will restore the state (image) of the database at the time of the checkpoint. That’s what the dump files are for; they record these transactions so that t
Related Questions
- Ingres Version: All OS: All Q: When I define the same location for checkpoint, journal and dump files Ingres gives a warning message. Is it legal to use the same location for these file types ?
- Ingres Version: All OS: All Q: What purpose do the dump files serve when I checkpoint my database?
- Ingres Version: All OS: All Q: How does an object get registered in star database?
