Subject | Re: [IB-Architect] GBAK processing |
---|---|
Author | Bill Karwin |
Post date | 2000-04-26T18:53:47Z |
How about this approach (assume databases are multi-file):
1. Once you finish backing up a database, restore it immediately to
different filenames than those used by the live database. You can do this
without interrupting the live database.
2. If the live database gets corrupted, rename its first file to something
else, just to get it out of the way.
3. Rename the _first_ file of the "hot spare" database to the same name that
was previously used by the first file of the live database.
4. Tell your users to reconnect.
5. Try to repair the corrupted database at your leisure.
The first file contains pathnames to secondary files. It's hard to move the
secondary files to different locations, but you don't need to in order to
connect to the database. You rename the _first_ file to the same name as
the first file of the live database.
Bill Karwin
1. Once you finish backing up a database, restore it immediately to
different filenames than those used by the live database. You can do this
without interrupting the live database.
2. If the live database gets corrupted, rename its first file to something
else, just to get it out of the way.
3. Rename the _first_ file of the "hot spare" database to the same name that
was previously used by the first file of the live database.
4. Tell your users to reconnect.
5. Try to repair the corrupted database at your leisure.
The first file contains pathnames to secondary files. It's hard to move the
secondary files to different locations, but you don't need to in order to
connect to the database. You rename the _first_ file to the same name as
the first file of the live database.
Bill Karwin