locked
Log Shipping with SQL Server 2008 RRS feed

  • Question

  • Hello !

    My environments are SQL 2008 STD x64 US SP1.

    On the principal site, I have 3 servers for mirroring (one of the 3 is a witness server) and on the DR Site, one server.With mirroring, there are no problem, all is ok, it runs ok. Databases are always reachable.

    If I choose Log Shipping architecture (for study), when the server transfers the informations from a site to another, is the database located in DR site , can be reachable directly ? What actions should be done ? Can I automate this (with a monitor or other) ? Is it possible to have an automatic failover with log shipping ?

     


    Thanks for advance for your ideas / help - Regards - Have a nice day ! RHUM2
    Wednesday, April 21, 2010 3:56 PM

Answers

  • Hi,

    U Dont have a automatic failover setup as provided by mirroring in log shipping however u can custom code it. Someting of the sorts which checks for the heart beat select @@servername in a job and then failover. However the redirection part of application would have to be done manuaaly.

     


    Apps Dev
    Wednesday, April 21, 2010 4:06 PM
  • Log shipping does not provide inbuilt automatic failover, you can write bespoke scripts to perform the Recovery action at the DR site to achive an unmanned failover.

     

    Wednesday, April 21, 2010 8:26 PM
  • I have this setup on 20 servers and the one thing you need to base the "automated failover" on is the business rules set in place for a true DR situation.  Failing over in HA is nothing like DR.  Typically your systems will need several other configuration changes, business processes changed for the new infrastructure etc...

    In that, you can create your own automated failover for log shipping using SQL Agent and monitoring the publisher of the logs.  In the mirroring landscape you can use the witness to run other jobs to actively validate the availability of the primary database.  If a failure is found, the first thing is to send the notifications out and then and only if the business allows it, fail things over.  This is very rare to be able to do for DR though.  

     

    As the others said, no native automated failover exists for LS as in most DR strategies.

     


    Ted Krueger Blog on lessthandot.com @onpnt on twitter
    Wednesday, April 21, 2010 8:37 PM
  • Hi RHUM2,

    If you are planning to go for mirroring then go through with below mentioned link it will help to provide broder way about mirroring.

    http://www.sqlservercentral.com/articles/SS2K5+-+High+Availability/3046/

     

    Cheers

    Praveen Barath

    Thursday, April 22, 2010 12:57 PM

All replies

  • Hi,

    U Dont have a automatic failover setup as provided by mirroring in log shipping however u can custom code it. Someting of the sorts which checks for the heart beat select @@servername in a job and then failover. However the redirection part of application would have to be done manuaaly.

     


    Apps Dev
    Wednesday, April 21, 2010 4:06 PM
  • Log shipping does not provide inbuilt automatic failover, you can write bespoke scripts to perform the Recovery action at the DR site to achive an unmanned failover.

     

    Wednesday, April 21, 2010 8:26 PM
  • I have this setup on 20 servers and the one thing you need to base the "automated failover" on is the business rules set in place for a true DR situation.  Failing over in HA is nothing like DR.  Typically your systems will need several other configuration changes, business processes changed for the new infrastructure etc...

    In that, you can create your own automated failover for log shipping using SQL Agent and monitoring the publisher of the logs.  In the mirroring landscape you can use the witness to run other jobs to actively validate the availability of the primary database.  If a failure is found, the first thing is to send the notifications out and then and only if the business allows it, fail things over.  This is very rare to be able to do for DR though.  

     

    As the others said, no native automated failover exists for LS as in most DR strategies.

     


    Ted Krueger Blog on lessthandot.com @onpnt on twitter
    Wednesday, April 21, 2010 8:37 PM
  • Hello !

    Thanks for these explanations. I think I perpetuate mirroring databases for my use.

     


    Thanks for advance for your ideas / help - Regards - Have a nice day ! RHUM2
    Thursday, April 22, 2010 8:15 AM
  •  

    Hello !

    Thanks for your response and these explanations. I think I perpetuate mirroring databases for my use.


    Thanks for advance for your ideas / help - Regards - Have a nice day ! RHUM2
    Thursday, April 22, 2010 8:16 AM
  • Hello !

    Thanks for these explanations. I think I perpetuate mirroring databases for my use.


    Thanks for advance for your ideas / help - Regards - Have a nice day ! RHUM2
    Thursday, April 22, 2010 8:17 AM
  • Hi RHUM2,

    If you are planning to go for mirroring then go through with below mentioned link it will help to provide broder way about mirroring.

    http://www.sqlservercentral.com/articles/SS2K5+-+High+Availability/3046/

     

    Cheers

    Praveen Barath

    Thursday, April 22, 2010 12:57 PM