Please start any new threads on our new site at https://forums.sqlteam.com. We've got lots of great SQL Server experts to answer whatever question you can come up with.

 All Forums
 SQL Server 2005 Forums
 High Availability (2005)
 Question

Author  Topic 

stumbling
Posting Yak Master

104 Posts

Posted - 2009-05-28 : 05:09:29
Hi there if i set up mirroring
and i have
a really good machine as the principal
a good machine as the mirror
an ok machine as witness
if the principal goes down for a week how can i get the really good machine to once again become the principal?
I have read http://technet.microsoft.com/en-au/library/cc917680.aspx and maybe i missed something but i could not see the way that the really good machine would be come the principal again.

Am i correct in assuming the following?
really good machine principal goes down
good machine mirror is promoted to principal
ok machine witness keeps going
if i bring really good machine back on line it now becomes the mirror correct?

So if this theory is correct then is this how it works
really good machine which is now the mirror stays on
good machine which is now the principal is turned off
ok machine witness keeps going
if i bring back the good machine it now becomes the mirror again correct or not?

Sorry for the long winded explanation just a bit confused.

Cheers

webfred
Master Smack Fu Yak Hacker

8781 Posts

Posted - 2009-05-28 : 08:07:26
I can not see in this moment, but I know there is a button to do a manual failover to change principal and mirror again.

Webfred


No, you're never too old to Yak'n'Roll if you're too young to die.
Go to Top of Page

mcrowley
Aged Yak Warrior

771 Posts

Posted - 2009-05-28 : 09:39:28
I would not intentionally bring a machine down to initiate a failover. You can use the ALTER DATABASE command to fail back to the really good machine when traffic is lower, and any impact to other applications can be minimized. One thing to note, you should check in sys.database_mirroring to make sure that the database is fully synchronized, before failing over. Otherwise, you may lose a bit of data.
ALTER DATABASE [name goes here] FAILOVER
Go to Top of Page

stumbling
Posting Yak Master

104 Posts

Posted - 2009-05-28 : 17:39:25
I would love to know where that is as i need to test the proccedure asap.
Cheers

quote:
Originally posted by webfred

I can not see in this moment, but I know there is a button to do a manual failover to change principal and mirror again.

Webfred


No, you're never too old to Yak'n'Roll if you're too young to die.

Go to Top of Page

stumbling
Posting Yak Master

104 Posts

Posted - 2009-05-28 : 17:41:51
So what i could do is on the good server check to confirm that they are synchronised then run the alter database command then the really good server would be promoted back to the primary again. After that i could start the good server back up and it would then take the roll of the mirror server again.

Correct?

Cheers

quote:
Originally posted by mcrowley

I would not intentionally bring a machine down to initiate a failover. You can use the ALTER DATABASE command to fail back to the really good machine when traffic is lower, and any impact to other applications can be minimized. One thing to note, you should check in sys.database_mirroring to make sure that the database is fully synchronized, before failing over. Otherwise, you may lose a bit of data.
ALTER DATABASE [name goes here] FAILOVER


Go to Top of Page

mcrowley
Aged Yak Warrior

771 Posts

Posted - 2009-05-29 : 09:41:29
Nope. No reboots are necessary. When you issue the ALTER DATABASE command, the primary becomes the mirror, and the mirror becomes the primary.
Go to Top of Page
   

- Advertisement -