Enterprise Storage Forum content and product recommendations are editorially independent. We may make money when you click on links to our partners. Learn More.
Data mirroring is a common approach to protecting data. It involves the copying of data in real time from one location to another, such as another local server or device or a remote storage medium as a facet of disaster recovery (DR). The mirrored data is an exact copy, so if one set of data is lost, the other is available.
Mirroring is often used to lessen risk. If an organization has all of its data in one location, it is open to risk. In the event of a loss of power, cyberattack, or disaster, having all data copied to another location can be a life saver. This approach avoids single points of failure within the infrastructure.
Another aspect of mirroring is that it is sometimes used in place of backups. As everything is automatically copied elsewhere in real time, some consider there is no need to continue backups. However, smart organizations employ both. Backups and mirroring should go hand in hand.
Here are the five key trends in data mirroring you should be aware of in your cybersecurity practices.
See more: Cloud Disaster Recovery Best Practices
1. Data Sharing
Data mirroring has many uses — DR, avoiding single points of failure, and retaining a second copy of data in real time. But a new use called data sharing is emerging.
“Despite being originally crafted to sidestep single points of failure in the storage infrastructure, data mirroring is now being applied in various data sharing scenarios,” said Augie Gonzalez, director of technical marketing at DataCore Software.
He offered the example of mirroring being used at the intersection of transaction processing and business intelligence (BI), where real-time online transactional processing (OLTP) data is mirrored to analytics tools. Those analytics tools can then delve into that data set to derive insight without interfering with the primary data set.
2. Data Proximity
The closer data and applications are to end users, the lower the latency will be.
If the primary site is on the other side of the country or half a world away, it can take a while for that data to travel across the network to where it is needed. Data mirroring is one way to reduce such latency.
“Data mirroring is gaining favor as a means of bringing data closer to the user in metropolitan area networks,” said Gonzalez with DataCore.
“Each site in a stretch cluster operates directly on the nearest mirrored copy rather than incur the delay of traversing LANs to reach the source image.”
3. Cloud Databases
Databases used to be retained solely on-premises. But these days, cloud databases are beginning to dominate.
With databases now in the cloud, mirroring provides much needed flexibility, latency, and agility.
“With cloud deployments, data mirroring takes on greater importance and particularly when core business applications, like analytics and databases, move to the cloud,” said Kirill Shoikhet, CTO of Excelero.
“Databases must have mirroring across availability zones. Technologies that can synchronously mirror, while minimizing the number of round trips have an advantage, as cross-availability zones and regions means that there is some basic latency due to the distance and physics.”
See more: Cloud Database Trends
4. Ransomware Protection
Ransomware attacks are happening with higher frequency, and they are rapidly growing in complexity. It’s getting tougher for many organizations to prevent them or even detect them in some cases.
Senior executives, therefore, are urging their teams to keep data highly available and secured against ransomware attacks to keep business running 24/7. This is doubly important in the remote model of operations that have emerged.
“To protect against ransomware, businesses will continue to replicate data into multiple immutable copies across both on-premises and multicloud systems with built-in air gap mechanisms, while the primary storage system provides data snapshot and versioning features to recover earlier versions from attacks,” said Param Kumarasamy, VP of product management at Commvault.
“Due to these trends, business leaders will be focusing less on on-premises data replication and instead ramping up the push for their application owners to adopt a multicloud strategy for high data availability for virtualization, enterprise, and cloud applications, by replicating the data across multiple cloud vendors.”
See more: Fortifying Your Backups from Ransomware
5. Serverless Replication
Replication used to be done between one server and another, one network attached storage (NAS) box and another, or one appliance and another.
However, that has changed as the cloud has become more pervasive. A trend noted by Kumarasamy with Commvault is serverless replication of data within and across the cloud. New era cloud applications have separated the data and compute layers, so each resource can be scaled independently.
“The data of these cloud applications will be replicated in multiple cloud zones within or across regions for high availability,” Kumarasamy said.
“Data availability is essential for most business-critical applications.”
