Protecting data is still an important requirement in the modern data center, but the role of backup software is changing. An enterprise-level backup software application faces three challenges:
- Modern primary storage has excellent native backup capability. As a result, it is reasonable to question the necessity of a separate backup process.
- User belief that backup software must be a once-a-night activity and that it will not capture data frequently enough to be of any value.
- The backup process creates a separate silo of storage that needs to be bought, maintained and managed.
Self-protecting primary storage
The primary challenger to backup software is primary storage itself. Modern primary storage systems have advanced RAID protection schemes that protect against multiple, simultaneous drive failures. The systems also provide advanced snapshots that can save an almost unlimited number of point-in-time copies of data. Snapshots once were vulnerable to a storage system failure, but modern storage systems can now replicate the snapshot data to a secondary system on site and a third system off site. If the primary storage system fails for any reason, users can go immediately to a second or third system without interrupting operations.
However, self-protecting storage has two weaknesses:
- It can be expensive. Without a backup offering, the second and third systems have to be the same size as the primary storage system, and they have to maintain all data forever. The data center is also paying to provide all applications and data a premium level of availability when the reality is that only a small percentage of data needs it.
- It lacks one of backup's biggest strengths: indexing. When backup software copies data, it creates and updates a database that makes it easier to find protected data.
Backup strikes back
The challenge of frequency stems from the belief that administrators can only execute backups once per day and typically late at night to avoid interfering with data center operations. Many modern backup systems have the ability to back up data at a block level, often referred to as Changed Block Tracking. Essentially, the backup software keeps track of what data changes at a block level and sends only changed blocks instead of the entire file. The result is a dramatic reduction in the size of each backup job by as much as 95%. In addition, most enterprise-level backup software application systems can leverage application or OS APIs to place databases or active applications in a backup mode and make quiesced copies. The combination of block-level incremental backups and clean application backups means data capture can be much more frequent -- several times a day or per hour.
The requirement of a secondary storage silo for enterprise-level backup software applications is actually an advantage because while modern primary storage systems are incredibly reliable and well protected, counting on a single system may put all your data eggs in one basket. Having an independent copy of data on a separate system protects against software corruption and outside cyberattacks.
Secondary storage systems can also replicate data to another secondary system in a DR site or to a storage pool in the cloud. Even if multiple primary storage systems are used to create a self-protecting pool of storage, leveraging backup software and secondary storage allows those second and third primary systems to be smaller in size. Essentially, they only need to be large enough to store data from the most mission-critical application.
Some modern backup offerings now present the data they are managing directly to other applications, a capability known as copy data management. Functions like test/dev and analytics all need access to data; instead of making copies on expensive primary storage, these processes can leverage the data found on secondary storage.
Finally, there is the indexing capability of an enterprise-level backup software application. Every file you copy into a backup application is tracked in a database. Most applications will track information like file name, owner, type, date modified and the number of versions protected. Some applications go a step further and provide context-level search of the data within the file. Primary storage systems, on the other hand, may have the ability to store information for a long period of time, but finding data on those systems is like finding a needle in a haystack.
Backup software is still relevant. It provides a secondary copy of data that is independent of the primary storage system. And it is no longer just an insurance policy; modern enterprise-level backup software applications can serve data via copy data management to help curtail the growth of primary storage while feeding important functions like test/dev and analytics. Backup hardware complements these advances by providing replication of data to secondary sites and cost-effective but solidly performing storage I/O for copy data functions.
Use Acronis backup software more efficiently in your data center
Discover how backup software can save your business money