Protecting against a data loss disaster

What a data loss disaster is  

A disaster is a situation in which service from one physical site (for example, a building or a corporate campus) on the network is lost for an extended period of time. The following are examples of disasters:

• Fire
• Earthquake
• Prolonged power outages at a site
• Prolonged loss of connectivity from clients to the filers at a site

What a disaster affects  

When a disaster occurs, it can affect all of your computing infrastructure including Network Appliance filers, application servers, networking connectivity, and client connectivity. When you create a disaster plan, take into consideration all of your computing infrastructure.

Determining if a disaster occurred  

It is critical that you follow some predefined methodical procedure to confirm that a disaster really has occurred. The procedure should determine the status of the supposed disaster site.

• Use external interfaces to the filer, such as the following:
  • Ping
  • Remote shell
  • FilerView
• Use network management tools to test connectivity to the disaster site.
• Physically inspect the disaster site, if possible.

You should declare a disaster only after determining that a disaster has occurred and that service cannot be restored.

Tools for protecting against data loss disasters  

The following Network Appliance features and products are best for enabling the administrator to back up or replicate data stored on a local filer to an off-site network location. This ensures data can be restored if data loss is caused by disaster at a primary data storage site.

Feature	Description
SnapVault Inter-site snapshot backup and restorability.	A SnapVault secondary storage device can be located offsite, any distance from the primary storage units that it is backing up. Data recoverability In event of a data-loss disaster at a primary storage site, data that is backed up to SnapVault secondary storage can be restored to primary storage units that have not suffered physical damage or that have been replaced either at the disaster site or at an alternate location. Currency of restore Data can be restored from the time that the last SnapVault snapshot was created on the secondary storage system. Connection requirements DSL connections or faster are recommended between the primary and secondary storage systems. Modem connections are possible. Routers, switches, and DNS servers should be pre-configured to direct users to alternate primary storage sites if the primary storage system that they first attempt to access becomes unavailable. Advantage Centralized, inexpensive off-site backup For more information See Chapter 5, Data Protection Using SnapVault.
SnapMirror Inter-site snapshot backup, availability, and restorability	A SnapMirror destination storage device can be located offsite, any distance from the source filer whose volumes it is mirroring. Data availability In event of a data-loss disaster at a source filer site, data that has been snapmirrored to the destination site can be made quickly available at the destination site. Data recoverability SnapMirror data can be restored to source storage units that have not suffered physical damage or that have been replaced either at the disaster site or at an alternate location. Currency of restore Data can be restored from the time of the last SnapMirror snapshot transfer from source to destination. Connection requirements DSL connections or faster are recommended between the source and destination filers. Modem connections are possible. Routers, switches, and DNS servers should be pre-configured to direct users to the destination storage site if the source filer they are attempting to access becomes unavailable. Advantage Combined off-site protection and availability For more information See Chapter 4, Data Protection Using SnapMirror.
Metro- Cluster Inter-site real-time backup, availability, and restorability	Synchronously mirrored MetroCluster filers can be located at different sites, up to ten miles from one another. Data availability In event of a data-loss disaster at one filer site, data that has been mirrored to the partner filer site can be made quickly available. Data recoverability This data can also be mirrored to source storage units that have not suffered physical damage or that have been replaced either at the disaster site or at an alternate location. Currency of restore Data can be restored from the time of the last NVRAM checkpoint. Connection requirements Data ONTAP cluster connections supplemented with switches and DSL or faster connections are required. Routers, switches, and DNS servers should be pre-configured to direct users to the MetroCluster partner if the clustered filer that they first attempt to access becomes unavailable. Advantage Combined real-time off-site protection and availability For more information See Appendix A, Disaster Protection Using MetroCluster.

Copyright © 1994-2003, Network Appliance, Inc. All rights reserved.
Part No. 210-00325
File: intro7.htm
Last updated: 01/14/03 17:50:30