SharePoint Protection and Recovery using DPM 2007 - Part II

In the previous post, we covered the steps necessary to setup DPM and SharePoint to allow recovery of any SharePoint object from a single file up to the entire Farm. In this post, we will cover the worst-case scenario in which the entire farm has to be restored. This is also the easiest of the recovery scenarios that we will be posting over the series.
The steps outlined below are based on the understanding that SharePoint and DPM are configured based on the information provided in the first post. Without further comment, let’s take a look at the steps necessary to recover a SharePoint farm whether the farm is available or not.

Recovering A SharePoint Farm

If the entire SharePoint farm must be restored, this can be restored over the top of the production servers in the farm. The search services may not be available for some 15-30 minutes, or longer depending on how large the infrastructure is, while the SharePoint farm’s contents are re-indexed following the restore.

Recovering a SharePoint Farm when the Farm is Available

In the event that a SharePoint Farm must be recovered while the Farm is still available for access, there are a few requirements however, that must be considered in order for the recovery to be successful.
First, the front-end web servers must be configured, as they were when the recovery point was created. If the configuration is different, this can cause the recovery to fail.
Second, the farm structure must be created on the front-end server as the farm data will be recovered to the existing structure.
Lastly, the SQL instances must be configured with the same names as they were when the recovery point was created.
Once the requirements have been met, the following steps will help to get the Farm recovered and returned to production.

1. From within the DPM Administrator Console, click on the Recovery tab.
2. In the Protected data: section, expand out the SharePoint server which contains the farm to be recovered.
3. Under the server name, click on the All Protected SharePoint Data entry. This will display the Farm name in the lower right-hand pane.
4. In the calendar display, select the date and time the Recovery Point was taken which is to be restored.
5. Under the Recoverable Item field, confirm that the farm displayed is the farm intended to be recovered.
6. Select the farm to be recovered under the Recoverable Item field by right-clicking on it. This displays the context sensitive menu with the options to Show all recovery points and Recover…
7. Choose the Recover… option which brings up the Review Recovery Selection window. Chose Next to continue.
8. In the event of a catastrophic crash, chose the Recover all Sharepoint content and components radio button. If you are restoring the farm for auditing or to satisfy some other requirement other than a full farm recovery to the original servers, choose the Copy all Sharepoint content and component files to a network folder. This allows the product data to remain intact while an older version is restored to an alternate location for auditing, etc. When you have made your selection, click Next to continue.
9. If you have SAN hardware that has the snapshot functionality enabled to clone and split a clone to make it writable and this SAN is connected to the DPM Server, then you can use this feature by selecting the Enable SAN based recovery using hardware snapshots checkbox in the Specify Recovery Options window. When you have made your selection, click Next to continue.
10. The Summary window is displayed and it lists in the Recovery items, Details: window all of the SharePoint files you will be recovering. The Total size of the restoration is provided as well. When you have made your selection, click the Recover button start the recovery.
After the recovery is complete, the search indexes will have to be rebuilt. Based on the size of the farm and the search configuration, this may take 15-30 minutes or longer. Once the farm is restored, users will be able to navigate its content immediately with some minor delay caused by the search indexing update.

Recovering a SharePoint Farm when the Farm is Unavailable

In the event that there is a catastrophic incident that causes the loss of hardware in addition to software, the entire SharePoint farm may need to be recovered to a new installation. When bringing the new hardware online, please note that the SQL Instance and the Front-End web server must have the same name as the servers that hosted the SharePoint farm when the recovery point was taken.

To recover the farm to an installation where the protected farm is not available, take the following actions.

1. Confirm the SQL Instance and front-end web server names match those at the time the Recovery Point was taken.
2. Install the necessary prerequisite software and the DPM Protection Agent on the front-end web server. Confirm the agent is communicating with the DPM Server.
3. Using either the stsadm –o registerwsswriter or the ConfigureSharePoint.exe (C:Program FilesMicrosoft Data Protection ManagerDPMbin) utility provided as part of the DPM Agent installation files.

Note: As a benefit of using the ConfigureSharePoint.exe tool, the steps in this section are completed automatically. This can be confirmed by acting on the steps above and noting that the account specified in the running of the utility is the account specified in the fields of the “Identity” tab.

4. From within the DPM Administrator Console, click on the Recovery tab.
5. In the Protected data: section, expand out the SharePoint server which contains the farm to be recovered.
6. Under the server name, click on the All Protected SharePoint Data entry. This will display the Farm name in the lower right-hand pane.
7. In the calendar display, select the date and time the Recovery Point was taken which is to be restored.
8. Now complete the wizard and allow the farm data to be restored.
9. Once the restore has completed, run the SharePoint Products and Technologies Configuration Wizard and disconnect the front-end web servers from the farm. All front-end servers must be disconnected.

Note: If the main front-end server for the farm is not the front-end server that DPM uses to protect the farm, you will also need to disconnect the front-end server that DPM uses to protect the farm.

10. Open the Internet Information Services (IIS) Manager console and delete all web site and application pool entries for the farm being restored.
11. Run the SharePoint Products and Technologies Configuration Wizard and select the option to connect to an existing SharePoint farm. Specify the server name and database name specified at the start of this process. These must be the same names as was in the Recovery Point when it was taken.
Note: Steps #10 and #11 must be completed on each front-end server in the farm.
12. On the Completing the SharePoint Products and Technologies Configuration Wizard page, click on Advanced Settings, and then click Next.
13. On the Advanced Settings page, select the option Use this machine to host the web site, and complete the wizard.
After the recovery is complete, the search indexes will have to be rebuilt. Based on the size of the farm and the search configuration, this may take 15-30 minutes or longer. Once the farm has been restored, users will be able to navigate its content immediately with some minor delay caused by the search indexing update.

Vic Reavis
Support Escalation Engineer
Microsoft Enterprise Platforms Support