Restore workflow

  • Just wanna make sure I have this straight in my head:

    doing full DB backup once/day, txnlog every 15 minutes.  Let's say disk device gets goobered up.  in my head, it should go like this:

    1- put DB in single user mode

    2- back up current txnlog to previsouly defined backup device (capture the tail)  truncate?  no truncate ??

    3- EntMgr > DB > all tasks > restore database

    4- options tab, choose " Leave database read-only and able to restore additional txn log.  this sould apply full plus periodic txnlogs except for "the tail"

    now, how do I apply the tail??

  • #1 - Not really necessary.  Just make sure there are no users when you start.  The restore itself will keep the users out. (suggest non-operational instead of read only.  read only allows users to connect).

    #2 - truncate? no truncate?  Who cares?  You're gonna restore over it!  Just don't do a TRUNCATE ONLY (which doesn't back it up) ! Plus, this is assuming that you have access to the database to back it up.

    #4 - again, non-operational instead of read-only.  to get the "tail", select "point in time restore" and specify a date/time to stop at.

    By the way, "read-only" places the database in standby mode.  This would generally be used in a log-shipping scenario where you want the users to be able to use the standby database for reporting.  AND, if you're backing up the tlog every 15 minutes, I wouldn't recommend allowing users in at all for log-shipping either!  I  am currently doing this and several times per day, the restore takes longer than 15 minutes, keeping the server busy restoring log backups.

    Steve

Viewing 2 posts - 1 through 1 (of 1 total)

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