If you're looking for a single backup solution, you're probably going to be disappointed. Backups serve multiple purposes within an enterprise: maintain the historical trail of data for compliance purposes, restore data if it's mistakenly deleted or corrupted, minimize downtime due to a device failure, and support rapid recovery when a disaster makes the primary data processing site unavailable.
Meeting those different needs means using multiple backup strategies. Here's a look at the options:
Replication makes a single duplicate copy of data available in near real-time. The main purpose of replication is to minimize downtime due to a device failure; processing simply switches over to using the replicated copy. There may be some data loss of records that weren't replicated to the secondary node before the failure. It's important to note that replication preserves only the current version of the data, meaning there's no historical trail and if data is deleted or corrupted, the replicated version reflects those changes.
Snapshots duplicate a datastore at a moment in time. These are generally very fast to create and also very fast and easy to use for data recovery; in fact, recovery procedures can be automated to be performed on user request. However, snapshots are typically saved on the same storage device as the original data. This means they can impact production performance. If snapshots are not application aware, they may capture inconsistent data. Another significant concern is that, as they are on the same storage device as the production data, if the storage device fails, you lose your backup as well as the production copy of the data. This concern can be mitigated by backing up the snapshot to a secondary storage device.
- Local backup
A local backup makes a copy of data and stores it on a separate device in the same data center as the original data. Because the data is stored locally on a distinct device, creating the backup is fast and it isn't affected if the production device fails and is easily accessed when needed. However, if there is a data center-wide outage, the backup storage device will also be impacted and unavailable.
- Offsite backup
Offsite backups, like local backups, make a complete, consistent copy of the data and stores it on a separate device from the production data. Unlike a local backup, offsite backups are stored at another location, either in the cloud or at a secondary data center. Using the cloud means data is immediately available, while secondary storage sites may not be fully staffed and there may be delays in accessing data. However, bandwidth issues may make restoring data from the cloud slow. In addition, because the cloud is pay-for-usage, it may not be the best location to store your historical data.
Creating an effective backup strategy requires evaluating your applications, data, business requirements, and compliance requirements in order to determine which level of backup and recovery support each data set needs. Once that's determined, you can evaluate the available backup software, backup appliances, cloud backups, and disaster recovery options in order to select the most cost-effective technology that will support those needs.
dcVAST offers a range of services to help companies use these different backup approaches effectively, including Managed NetBackup, Managed Enterprise Vault, and Disaster Recovery as a Service. Contact us to start developing and implementing a backup strategy that addresses all the reasons you need data backups.