You are a database developer and you have about two years experience in creating
business intelligence (BI) solutions by using SQL Server2008. Now you are employed in a
company which is named NaproStar and work as a developer of a SQL Server 2008
Integration Services (SSIS) for the company. To meet the business requirements, SSIS
should be used on a SQL Server 2008 failover cluster. There is a complex package on your
server and it needs a long time to start it. Now your manager asks you to make sure that the
package can be resumed when it fails to execute the failover. What action should you
perform?
A.
First you should have checkpoints implemented within your package, and then restart the
package whenever a failover occurs
B.
First you should have transactions implemented within your package, and then restart the
package whenever a failover occurs.
C.
First you should cluster the SSIS service, and then restart the package whenever a
failover occurs
D.
First you should cluster the SSIS service, and then have the SSIS service added to its
own cluster resource group
E.
First you should cluster the SSIS service, and then have the SSIS service added to the
SQL Server cluster resource group.
Explanation:
EXAM TIP
When you are running a package on a server node of a Windows cluster environment and
the node fails, the restartability rules apply. you can turn on checkpoints in your packages
and have the checkpoint fi le created on a share so that if the package needs to be
restarted, it can locate and use the checkpoint file.
Simply put, there are two ways to configure the SSIS service in a clustered environment:
* installing the ssis service independent from the cluster resources You can install the SSIS
components on all nodes of the cluster, and by default, they will not be part of any cluster
resource group. The service will be started on all nodes, and you can run packages from any
node in the cluster. If all your packages are stored on a network share that is not part of the
package store, no further confi guration is needed. If you want to centralize all your
packages in SQL Server, you need to change the MsDtsSrvr.ini.xml fi le. Change the <
ServerName> element to reference a specific server and instance; if the SQL Server is in
the cluster, use the virtual server name and instance. Last, change this fi le on all the nodes
so that you can connect to the SSIS service on any machine and see the same packages.
* integrating the ssis service in a cluster group You can add the SSIS service as a cluster
resource, a process detailed in the “Confi guring Integration Services in a Cluster
Environment” white paper mentioned earlier. If you want to store packages in the package
store, you would choose this approach because the shared service would be running on only
one node at a time and you could reference the virtual name of the server. The service
would need to point to a shared MsDtsSrvr.ini.xml fi le on a shared drive in the same cluster
resource group as the service. This requires a registry change, which is also documented.
The package store location also must be on the shared drive in the same cluster resource
group.