Save Your Data from Vanishing with Disaster Recovery Plans
The database server is cleaned up, tuned up, in top shape. How do you protect all that work, and the data contained in it, when a hurricane hits? A disaster recovery plan.
Disaster recovery is basically a process of dealing with failures. Failures like a server crash, theft, or a natural disaster. Plans are put in place in case they’re needed, with the hope they won't be.
Common disaster recovery technology includes clustering, database backups, log shipping and replication.
Clustering allows a database to "failover" (switch operation) to another server in minutes. Great for hardware failures; not so much for natural disasters (clustered servers are physically close together). The best way to resume operations after an earthquake or hurricane, for example, is to use an offsite recovery plan. Plans are built based on budget and business needs. Two common methods are installing a server in a remote office or contracting with a storage vendor.
Budget for Recovery – But Don't Break the Bank
Your database engineer/consultant will recommend which systems you should preserve, how, and where, based on your company's priorities. Cost, for instance is one of offsite recovery’s biggest factors. Smaller businesses can’t justify a database cluster on-site, or a dedicated remote server for offsite recovery.
But they're no less affected by a disaster, and will need recovery. The answer's in the planning. Small business plans – indeed, any disaster recovery plan – should include the following precautions:
- Replicate important data, and ship it to another location. Tape backups work well for the cost-conscious.
- Document your network’s structure, database tables, warranty numbers and company procedures. This way you can “restart” operations elsewhere.
- Note where hardware comes from within the documentation, including manufacturer/reseller contact information.
- Store software CDs or copies at the other location, for reinstallation of key systems.
- Craft a plan on who should do what in case of emergency. Talk with your database engineer/consultant for recommendations.
- Test your recovery plan regularly (once a quarter is recommended), so you’re sure it works. Make changes as necessary.
Save Your Data, Save Your Company
As you can see, not everything in a disaster recovery involves a database. But data preservation is the most important point. And a database server makes an ideal piece to build your plan around.
In Article 5 we'll go back through all the major points in our Health Check Series, and talk more about the value well-maintained database servers give. Also, a few questions to ask to determine if your DBA really knows their stuff.
Questions? Email our DBA team at info@planetmagpie.com. Or call us at 408-341-8770 to schedule an appointment.
