Post reply

SQL Server "Recovery Scenarios" Document

  • May 19, 2006 at 12:19 pm

    #113036

    I've worked in both the SQL Server and the Oracle environments for a number of years now.  One thing that I have found unique to Oracle is the considerable availability of information on what is referred to as "Recovery Scenarios."  In these documents, such as the one found here: http://www.uaex.edu/srea/bkupreco.htm go into great detail on all the different things that might break in Oracle and what actions should be taken to restore them back to their operating condition.  I've found some discussions of individual scenarios relative to SQL Server, such as how to recover from a corrupted master db, or a complete rebuild of an instance along with some very nice scripts, but I have not found a comprehensive document, which enumerates the vast majority of the problems that might occur & details on how to fix them.

    So, my question is, does anyone have or does anyone know where I might find such a document?

    Thanks!

  • May 22, 2006 at 8:00 am

    This was removed by the editor as SPAM

  • May 25, 2006 at 12:52 pm

    #640149

    Inside SQL Server is a book that would give you good information. Books Online is another resource.But if you are looking for scenarios when recovery would be needed, I guess you would have to go by examples in each of the topics.

  • May 25, 2006 at 12:58 pm

    #640153

    So, I gave this a shot, any feedback would be appreciated:

    - Complete loss of O/S and SQL Server binaries (their single solution suggests that this is what will happen with each and every failure);
    - Loss or corruption of master database (which uses the Simple Recovery Model);
    - Loss or corruption of any other databases (which generally use the Full Recovery Model);
    - Loss of transaction files (the resolution of this scenario will be influenced by the recovery model used);
    - Loss of disk containing backup dump files, prior to tape backup;
    - Failure of 'Restore VerifyOnly' operation of backup dump files (appears to not currently be done on after backups);
    - Loss of 'Read Only' data files/file groups (I don't believe any are defined in this environment, but a verification is needed);
    - DBCC CHECKDB failure or errors w/REPAIR_ALLOW_DATA_LOSS parameter specified;
    - Failed database recovery (what to do if the recovery process itself fails);
    - Replicated database recovery (special consideration is needed here, if there is any replication); and
    - Log shipping disruption or failure.
     

    I'm thinking that I need to create a process or decision tree that would allow a less experienced administrator to determine what scenario they are faced with and actions to take to recover.

Viewing 4 posts - 1 through 3 (of 3 total)

You must be logged in to reply to this topic. Login to reply