Mikhail Khomenko · May 12, 2016

Add database to mirror on Backup Failover node

Good day, Cache developers and users!

At first, sorry for my bad English -)  I have a little question.

As I know, we can add database to mirror only on Primary Failover Node, using SMP or ^MIRROR. And the usual procedure after that is to copy mirrored database to Backup Failover Node and restore it there (by Backup restore or by Activate-Synchronize procedure). This all works fine. But what if I first have copied the database from Primary to Backup and only than added the database to mirror on Primary Failover Node? (human forgetfulness). If I have a big database, it's not convenient to again copy it from Primary to Backup. As I think (I could be wrong), the adding procedure just rewrite some bytes in the beginning of database, for example, save mirror GUID and possibly save the last journal entry point. So if there is a method to add database to mirror on Backup Node? Of course, after that we need to restore the recent changes from Primary journal files on Backup Node.


0 378
Discussion (2)1
Log in or sign up to continue

There is no utility to do this.  You're right that to create such a mechanism is just a matter of manipulating the right bits and bytes just so, but it does mean that you'd lose the guarantee that these are identical copies, so we haven't created one.  The only context in which anything like this is available is the special case of converting shadow systems to mirror systems where we do have a migration utility that doesn't require completely resynchronizing the databases.

Thank you, Ray! Something like this I've supposed. Ok, we'll be more careful in the future.