May 26, 2008 at 10:10 am
In log shipping if the primary server fails then we can make the secondary server running.
Then how the client application will be redirected to the secondaruy server.Is there any standard procedure for yhis to happen
May 26, 2008 at 4:14 pm
It will depend on how your application is configured to access the database server.
If you use ODBC DSNs, then you may need to reconfigure the DSN per computer.
If the application stores the database server name in a config file or similar, you may need to change it there.
You could also edit the DNS entry for the primary server and change the IP address to match that of the secondary server, so anything looking up the primary server name will get directed to the secondary server. This could cause issues if you are rebuilding the primary server, however. Also, client machines would probably have the original IP cached, so some per-client work would still need to be done.
Presumably your SQL logins/database users will be matching up OK on the secondary server, otherwise things won't work properly.
MARCUS. Why dost thou laugh? It fits not with this hour.
TITUS. Why, I have not another tear to shed;
--Titus Andronicus, William Shakespeare
May 27, 2008 at 2:59 am
If you are using ODBC, on user network login you could run a registry script to modify the ODBC connection to point at the new server for all your users. There was a previous discussion on how to deploy ODBC by script on this web site, see
http://www.sqlservercentral.com/Forums/Topic456846-146-1.aspx
Or some applications allow you to configure multiple ODBC connections for different servers. If your application allows you to do this then you can configure ODBC connections to both your servers. Then if the first fallsover you can notify your users by mail/network login message to connect to the "yyyy" database instead of the normal "xxxx" database.
May 27, 2008 at 8:45 am
One thing is certain, Manual intervention is needed to get a log shipped secondary server database up and running as the primary database.
Gethyn Elliswww.gethynellis.com
May 27, 2008 at 11:03 am
It's depend on the Configuration for the DR Solution. Why we need DR? what is the baseline for that? How you can configure for that?
DR means if primary fail then DR will become Primary and take effects of the Primary server.
MCP, MCTS (GDBA/EDA)
May 29, 2008 at 7:02 am
Every application has there DR TEST Plan. Before actually go to DR Application will do the DR Test on the Production and DR.
It will be possible for all the application whoever use the DR server. Application will do manually failover to DR and connect there application with to make sure DR is in PLACE and working fine. Then again they will move to PRODUCTION.
MCP, MCTS (GDBA/EDA)
Viewing 7 posts - 1 through 6 (of 6 total)
You must be logged in to reply to this topic. Login to reply