Recovering a failed filer and server to a replicated environment

Ensure that you have rsh access to the filer hosting the SnapMirror destination volumes.

For more information about managing access from administration hosts, see the System Administration Storage Management Guide for your version of Data ONTAP.

Ensure that the filer hosting the SnapMirror destination volumes is online and that the volumes are available and shared through CIFS.

For information about creating CIFS shares, see your Data ONTAP System Administration File Access Management Guide. For information about general filer administration, see your Data ONTAP System Administration Storage Management Guide.

Follow the steps in the appropriate Microsoft Knowledge Base article to reset the computer account and rename the standby Exchange server in your new environment to the name of the original Exchange server:

• If your original Exchange server was running Exchange 2000, use KB article 297289.
• If your original Exchange server was running Exchange 2003, use KB article 822945.
• If your original Exchange server was running Exchange 5.5, simply perform an account reset.

  Caution
  Do not install Exchange yet. This is done in a later step.
  
  Make sure that you perform an account reset, rather than deleting the account and re-creating it.

Make sure that the same version of SnapDrive that was running on your original Exchange server is installed on your standby Exchange server.

From the MMC on the standby Exchange server, use SnapDrive to connect to the LUNs in the active file system of the SnapMirror destination volume. Make sure that you use the same drive letters for the same LUNs as your original configuration.

For information about connecting LUNs, see your SnapDrive documentation.

Note
If you get an error when connecting, such as "Disk is in exclusive use," use FilerView to ensure that the LUN is unmapped from all igroups and online.

Result

The mirror is broken and a Single File Snap Restore (SFSR) is performed using the most recent SnapDrive snapshot.

If...

Then...

Your original Exchange Server was running Exchange 2000 or 2003

Using the Microsoft Knowledge Base article you used in Step 3, install Exchange and any required Exchange service packs in disaster recovery mode on your standby Exchange server.

Result

When you complete this step, your standby Exchange server should be running exactly the same version of Exchange, including service packs, as your original Exchange server.

Your original Exchange Server was running Exchange 5.5

Install Exchange 5.5, using the same site and organization names as your original Exchange server, and then apply any required service packs.

Result

When you complete this step, your standby Exchange server should be running exactly the same version of Exchange, including service packs, as your original Exchange server.

On the standby Exchange server, install the same version of SnapManager as was running on your original Exchange server.

Launch SnapManager and run the SnapManager Configuration Wizard to ensure that the correct configuration is used.

If necessary, modify the configuration so that it matches exactly the configuration before the failure.

Restore the most recent backup using SnapManager Restore.

For more information about using SnapManager Restore, see Restoring Databases with SnapManager Restore.

Launch the Outlook client and verify that your data was properly restored and that the standby Exchange server, which now becomes your primary Exchange server, is operating correctly.

After you repair or replace your primary site, you can return to your original configuration. For more information about returning to your original site replication configuration, see Returning to your original site replication configuration.

Prepare primary site

Repair or replace the filer on your primary site.

For more information about setting up your filer, see the System Administration Storage Management Guide for your version of Data ONTAP.

Reconstruct the Windows environment, including the Active Directory Domain Controller, on your primary site by replicating the environment back from the replicated site.

Create temporary mirror

Using SnapMirror, create a mirror with the replicated site as the (temporary) source and the primary site as the (temporary) destination.

Note
Remember that a destination volume must be at least as large as its source volume. If your original destination volume was larger then your original source volume, you need to increase the size of the temporary destination volume to match the size of the temporary source volume.

If your original primary filer was not affected by the disaster, you can use the snapmirror resync command to perform this step more quickly. For more information about this command, see the System Administration Data Protection Guide for your version of Data ONTAP.

Capture Exchange data

On the standby Exchange server, perform a SnapManager backup and verification of all storage groups.

Block all user access to Exchange.

Note
This step is necessary because all changes to the Exchange data store made after the mirror is updated are lost when you switch over to the primary site.

Use SnapDrive to perform an update mirror operation using the following sdcli command once for every volume containing Exchange or SnapInfo data:

sdcli snap update_mirror -d drive

For more information about the SnapDrive command-line interface, see your SnapDrive documentation.

Shut down the standby Exchange server.

Reconstruct primary Exchange server

If...

Then...

Your original Exchange server was running Exchange 2000 or 2003

Using the appropriate Microsoft Knowledge Base Article for your version of Exchange, reset the machine account in the domain.

Caution
Make sure that you perform the account reset, as specified in the Knowledge Base article, rather than deleting the account and re-creating it.

Your original Exchange server was running Exchange 5.5

Rename the primary Exchange server to its original name.

Install and configure Windows on the primary Exchange server.

Caution
If you repaired the original Exchange server, wipe its disks and reinstall the operating system. Do not attempt to bring the server back online with the original software.

Install any Windows service packs that were installed on the primary Exchange server.

Ensure that you have rsh access to the newly reconstructed or repaired filer.

Install and configure SnapDrive on the primary Exchange server.

Using SnapDrive, connect to the LUNs in the active file system of the newly reconstructed or repaired filer.

Caution
Be sure you use the same drive letters that were used in your original primary site, and as should be in use in your recovery site.

Result

The reverse direction mirror you created in Step 3 is broken and an SFSR is performed.

If...

Then...

Your original Exchange server was running Exchange 2000 or 2003

Using the appropriate Microsoft Knowledge Base Article for your version of Exchange, install Exchange and any required Exchange service packs in disaster recovery mode on the primary Exchange server.

Your original Exchange server was running Exchange 5.5

Install Exchange and any required service packs on the primary Exchange server, using the same site and organization names as your original configuration.

Install the same version of SnapManager for Microsoft Exchange that was installed on the primary Exchange server.

Launch SnapManager and run the SnapManager Configuration Wizard, making sure to keep all data on the same drives where it was before.

Restore and verify Exchange data

Use SnapManager Restore to restore the backup you created in Step 4.

Launch the Outlook client and verify that your data was properly restored and that the new primary Exchange server is operating correctly.

Re-create mirror

Delete the (reversed) mirror configuration you created in Step 3.

Re-create your original mirror, with your newly reconstructed filer as the source, and the replicated site as the destination.

Reconfigure replicated site

Return your standby Exchange server to standby mode by performing these steps:

1. Wipe the disks.
2. Using a different name than the primary Exchange server, reinstall the correct version of the operating system and any required service packs.
3. Place the standby Exchange server into a workgroup. (This step is optional.)

Result:

Your replicated site is ready once again in case of the loss of or damage to your primary site.