Snapshots provide near-instantaneous data protection. They achieve this via point-in-time copies that focus on the storage metadata rather than on the data itself. As such, they record where each block of data is stored. Because this doesn’t take up much space, they can be made quickly – unlike backups that can take many hours as every bit of data must be transferred across the network.

Here are five use cases where snapshots can and should be used for databases and storage volumes.

1. Protecting data from ransomware

Jerry Rozeman, an analyst at Gartner, promotes snapshots as an additional safeguard against ransomware. Taking regular snapshots of databases and storage volumes offers a last line of defense if other cybersecurity protections are breached.

“Unstructured data platforms like network-attached storage, scale-out file systems and object storage provide inadequate protection from malicious deletion, encryption and data exfiltration, making them an easy-to-attack target,” said Rozeman.

If intelligent file indexing and immutability are included, the level of protection is further increased. Indexes of all files deliver version tracking and recoverability of files across snapshots. In addition, immutable snapshot features provided by some vendors can integrate with the detection and recovery tools used for ransomware detection.

Pure Storage, for example, provides immutability via SafeMode, a feature included in its Purity operating system for running all-flash arrays. By activating SafeMode, snapshots cannot be deleted by anyone. Thus, if hackers infiltrate the network, they may cause damage, but the snapshots will remain unharmed.

“Using SafeMode, periodic snapshots of volumes offer protection against rogue administrators, compromised credentials or attempts by ransomware gangs to delete backups and snapshots,” said Anthony Nocentino, principal field solutions architect at Pure Storage. “SafeMode makes it impossible to delete data snapshots.”

2. Toward zero recovery-point objectives

Organizations try to balance the ideal of a zero recovery-point objective (RPO) with the reality of achieving it in terms of financial outlay. When an outage, data breach or disaster happens, it is common for system recovery to only be able to roll things back to a day ago or sometimes as much as a week or more ago. The backups and other safeguards in place can only recover data up to that point and no further. Anything later is lost during the event. But in transactional applications or financial databases, such losses could add up to millions of dollars in business.

“The ideal recovery-point objective is zero,” said Jason Lohrey, CEO of Arcitecta. “Anything less is a compromise.”

Here is the logic: Backups typically run every few hours or perhaps once a day, meaning the RPO could be hours to days. Snapshots are much better and can be run hourly or even more often if the business case is there. And those with the resources can take things a step further by using synchronous replication and other technologies to ensure that not a single transaction is lost in the case of an outage.

3. Database restores

Nocentino noted that some databases and datasets are so large that it takes ages to restore them from backups.

“Split your databases up into segments, or arrange it so Oracle RMAN, SQL and MySQL are backed up and snapshotted separately so you can recover fast,” he said.

This also provides a solution in the event that data loss or failure occurs in only one of the databases in use in the enterprise, or within one segment of a larger database. If that is the case, the snapshot of that segment or section can be used for a very rapid restore.

In addition, Nocentino suggested relying on snapshots more than backups for primary or top-tier storage as they offer instant point-in-time recovery. Snapshots can serve the organization well for small amounts of mission-critical and hot data. For everything else, backups are fine. This approach ensures fast recovery and access to the data that is most urgently needed with standard backups relied upon to recover everything else.

That said, the traditional 3-2-1 system still applies. Retain three copies of your data, which means a primary copy as well as two additional copies saved on two different media, and with one copy retained offsite.

4. Moving large amounts of data

Data migration is a constant fact of life in modern IT. Organizations are shifting databases from on-premises to the cloud (and back again) or from cloud to cloud. They also migrate data from one application or vendor platform to another. Backups can be used for this. But when the database is too large or the storage volume colossal, it is clunky and absorbs enormous amounts of time and network resources. It can also be expensive when you factor in cloud egress fees. Using snapshots instead can take the time out of such operations.

5. Application development

Application development is a dynamic field. Developers are constantly spinning up new VMs or containers, dumping large databases into them and testing them to ensure all is working well. Sometimes things go wrong. When the developer wants to return things to a previous state, snapshots come in handy. Similarly, developers may move data to the cloud for testing and want to move it back on-premises on successful conclusion (or vice versa), as another example. Snapshots offer developers a tool that simplifies the data movement process.