Today’s digitally transformed enterprises operate in the moment and simply cannot afford downtime, even if it’s just a few moments. For that reason, many companies are developing strategies and approaches that ensure that the data they need—or customers are viewing—will always be available, regardless of what may be happening behind the scenes.
Keeping data highly available is no longer a simple matter of backing up one database with an up-to-the-second copy of another. Today’s enterprise data environments are complex, with many database platforms and modes of computing. A hybrid data infrastructure calls for new approaches to data replication.
This is one of the takeaways of a recent survey of 285 IT managers and professionals, conducted by Unisphere Research among members of the Independent Oracle Users Group. The study was conducted in partnership with Quest Software, representing a fairly broad sample of company types and sizes (“Achieving Your 2018 Database Goals Through Replication: Real-World Market Insights and Best Practices: 2018 IOUG Special Report on Database Replication Trends,” December 2017). Just about every organization has turned to real-time database replication as an essential inclusion within its availability and continuity architecture stack.
The survey found that databases continue to expand in size and complexity, while at the same time, more enterprises are turning to cloud-based resources to keep information highly available. The survey finds agreement that availability and performance are the most important services data shops can deliver. Big data keeps getting bigger and is growing faster by the year.
Most Important Database Concerns (rated as “Important” to “Extremely Critical”)
Availability of applications (no downtime) 63%
Database security 35%
The survey also examined the depth to which cloud computing is being adopted. Cloud is a key factor—and enabler—of enterprise data replication strategies going forward. While most major enterprise databases are still on-premise, this is likely to change soon. In 5 years’ time, a majority of respondents expect to have their major databases running in cloud environments.
Databases that Support Business-Critical Workloads: Oracle
Now In 5 Years
Oracle on premise 59% 25%
Oracle in the cloud 15% 52%
Databases that Support Business-Critical Workloads: SQL Server
Now In 5 Years
SQL Server on premise 70% 24%
SQL Server in the cloud 13% 51%
Data replication is considered a vital solution that supports infrastructure and data services by just about all enterprises, the survey shows. Eighty-nine percent reported employing data replication to keep vital data available to users and customers. Disaster recovery is the most visible use case, but offloading non-critical workloads and supporting data integration also are facilitated through replication.
When asked about the business drivers for supporting a robust replication initiative, disaster recovery—and by extension, business continuity—emerges as the leading factor, cited by 83%. While disaster recovery was a process that lasted days, then hours, then minutes, these days, an effective DR is a matter of split-second failovers to backup sites, clouds, and datasets, in which end users will likely never notice something was amiss. Disaster recovery is a great use case for data replication, but there are other applications as well.
Where, How, and Why Database Replication Technology is Employed
Disaster recovery 83%
Offload reporting workloads from transactional databases 49%
Data integration 44%
Offload backup workload from transactional databases 41%
Peer-to-peer replication 33%
Distributing data across enterprise 20%
Close to half of data managers also note that robust replication ensures the seamless offloading of reporting workloads from overloaded transaction databases. Data replication is about minimizing downtime and having data immediately available, versus the role of traditional backup which is all about redundancy. It’s important that data managers and their business counterparts understand these distinctions.