• Write down a disaster recovery plan. Make sure the plan and any other resources needed to perform a recovery are stored properly, and that staff know where they are.
• Keep the plan and other resources current. Make this step part of your regular checklist when making changes to the hardware or software configuration on your servers.
• Make sure more than one person knows how to execute the disaster recovery plan.
• Any disk sets that are created for recovery purposes should be duplicated, verified, locked, and copies stored both on- and off-site.
• Keep an eye on your disk hardware: run drive and controller diagnostics on a regular basis, listen to what NetWare tells you, and don't get caught without replacement hardware in the event of a drive or controller failure.
• If you don't have a contract for server or hardware swaps, make sure you have a good relationship with a local or catalog supplier who can get you the parts you need when you need them. If you can afford to have spare drives and other hardware on hand that's great, but not very practical. Don't rely on being able to "borrow" hardware from your fellow admins either - just like you, their ability to keep hardware on hand without it being in use is limited.

DOCUMENT

• Keep your server log current. Having details on hardware, firmware, and software/patch information in your files will make it much easier to get technical support and/or replacement parts when you need them.
• Use CONFIG.NLM, CONFGNUT.NLM (found in.NetWare.ITD.UMich:SHARED\TABND2\DIAGS\CONFIG) or a similar product provided with your backup software to take snapshots of your server configuration on a regular basis. Make this step part of your regular checklist when making changes to the hardware or software configuration on your servers. Don't leave this information only a network drive - archive these snapshots locally, on floppies, or as printouts with your server logs and disaster recovery plan.

BACK UP

• Read, understand, and follow all the documentation that comes with your backup software about planning and preparing for disaster recovery.
• Follow the advice of your manuals or software on media rotation, archives, and on- and off-site media storage.
• Keep an eye on your backup hardware: perform recommended cleanings; run tape drive diagnostics on a regular basis.
• Test your backups after any change to hardware or software and periodically in between. Test both partial and full restores.

COMMUNICATE

• Be prepared to communicate with your users. As much as possible, have prepared texts ready before you find yourself in a recovery situation, so you can quickly inform the users of the problems, issue an estimate for return of services, and concentrate on the recovery. Keep this communication going during the recovery process - your users will appreciate it.
• Be prepared to communicate with the NetWare support staff and other admins. You may need help or advice with NDS or other issues, and keeping the netware.support@umich.edu group informed is vital. The netware.administrators@umich.edu list is another useful place to gain from the experience of others.

RECOVER

• Contact netware.support@umich.edu so that they can help you by removing remnants of your failed server from the Michigan tree, if necessary, before beginning your recovery process.
• If you need to do a brand new install, be sure to apply all patches that were installed before the failure. This applies to both core OS patches and patches for any other software running on the server.