Data & System Availability
Data & System Availability - Backup
| Article Index |
|---|
| Data & System Availability |
| Storage options |
| Indexing services |
| Virtualization |
| Failover sites |
| System Portability |
| Backup |
| More information |
| All Pages |
Backup
It makes sense to have the backup offsite from the main site. In case of loss of the main datacenter, the data will still be available somewhere else. Having the backup offsite does not qualify as a disaster recovery solution since there are no servers at the recovery site to deliver the data to clients and applications. Also, there is usually no infrastructure present at a site that just delivers backup services. Building a production environment at such a site is often physically impossible.
Creating a traditional backup requires an agent inside the machine. It doesn't matter if this is a physical or a virtual machine. From this agent, data is sent directly to the backup server over the LAN. When these servers are virtual, it's also possible to create a backup straight from the virtual hard disk that resides on the SAN. Creating a backup this way leaves all resources available for virtual machines to run production instead of needing them to run backups. The disadvantage of this method is that it takes a snapshot of the data without application awareness so the data that's being backed up is in an inconsistent state. This method therefore only works for crash recoverable services.
Figure 12: Proxy VM backup