High Availability with SQL Server 2000 Failover and DoubleTake Part III

In my previous
article
, we discussed how to configure and test a typical failover scenario
using Double-Take. At this point, you should now be able to create a
replication set and configure replication and mirroring. In the final article
of this series, we will show you how to restore your SQL Server data once a
failover has occurred.

Case in Point

In this article, we will assume that you are running two
production SQL Servers with Double-Take software loaded for High Availability
and Disaster Recovery. At some point during the night, you had either a network
outage, power outage, disk failure, and/or a combination of problems. The
Double-Take software failed over to your Target SQL Server and you had no
downtime. You were paged last night about the failover and you made sure
Double-Take was performing as promised. Everything was working soundly so you
slept easily, or not, and were prepared to bring the Source server back online
in the morning.

Restoring your SQL Server 2000 Data

Your first step is to make sure that your original source
server is not connected to the network. Once you have disconnected it from the
network, resolve the issue that caused the problem. Once the issue is resolved,
you are ready to restore your SQL Server data.

Now that your issue is resolved, please make sure the
following services are shut down on the source server: DTC, MSSQLSERVER,
SQLSERVERAGENT, and Message Queuing. Now, select Start | Programs |
Double-Take |Failover Control, as shown in Figure A., on the target machine that is
acting on behalf of the failed source server.



Figure A:
The Failover Control Center displays that Belle has failed
over to Beast and has taken over Belles IP address.

Now, select the failed source, as shown in Figure A, and
click Failback. The prefailback script discussed in High
Availability with SQL Server 2000 failover and Double-Take-Part II
stops
the applicable services so no additional changes can be made. Next, you will be
prompted if you want to continue monitoring as shown in Figure B. Leave this dialog box open and we
will choose the appropriate option later in the process.



Figure B:

During the Double-Take restore process, do not choose to
Continue or Stop at this point in the progression.

Now you are ready to restore the data and your first step is
to connect the original source server back to the network and test
connectivity. Next, open the Double-Take Management console from the Start Menu
(Start | Programs | Double-Take, Management Console) and highlight the source
machine to make sure the replication set is disconnected. Simply right-click on
the replication set and choose disconnect.

In order to start the restoration process, choose Tools |
Restoration Manager from the Double-Take Management Console as shown in Figure C.




Figure C.

The Restoration Manager allows you to restore data back to
your original source.

On this window, the process is as follows: Place the
original source server name prior to the failover in the Original Source text
box. Next, click the dropdown and select the name of the server you want to
Restore From. This will be the target machine that took over for the failed
source. Now, choose your Replication Set name and then choose the name of the
machine where the data will be restored. Once you have entered the appropriate
information, click Restore.

Note: The time required for the restore depends on the size
of data being transferred.

When your restoration is complete, restart the SQL Services
on your source machine and reestablish your replication set. Finally, your data
has now been restored and you have successfully restored your data back to your
original source machine.

Note: Remember to click Continue now as shown in Figure B to
reestablish your Failover Monitoring.

»


See All Articles by Columnist
Steven S. Warren

Steven Warren
Steven Warren
Steven S. Warren is a popular author residing in Winter Haven, Florida with his wife Danna and 2 children: Catie-Charlotte and Dain. As a columnist on such well-known IT web sites as Techrepublic.com, CNET, and ZDNET, Steven has published numerous articles. Additionally, Steven holds the following certifications: MCDBA, MCSE, MCSA, CCA, CIW-SA, CIW-MA, Network+, and I-Net+. As a Senior Technical Consultant for The Ultimate Software Group, Steven has become an expert at administering Microsoft networks including Microsoft SQL Server. He is also a computer hardware and troubleshooting expert, and is constantly seeking out new technologies and certifications. Additionally, Microsoft recently awarded him the Most Valuable Professional (MVP) award for his outstanding achievements. Steven resides in Winter Haven, Fl.

Latest Articles