- Cloud migration is not a one-way-street, according to a Thursday report commissioned by Veeam, a data backup, recovery and management software company.
- The vast majority, nearly 9 in 10, of more than 1,700 IT leaders surveyed said their companies brought cloud workloads back to their data center for one or more reasons, including development, cost, performance optimization and disaster recovery.
- Shifting workloads multiple times between data centers and public and private clouds complicates security. It can make it harder to recover from data breaches, ransomware and cyber incursions, according to the report.
Data may be the substance fueling the latest gold rush, but it’s far more ephemeral — and harder to protect — than your average precious metal.
Most companies take measures to protect against breaches and plan for rapid recovery. But tracking workloads can be difficult in a hybrid cloud ecosystem where data can move from on-prem servers to multiple cloud platforms.
Nearly half of respondents said their company had brought applications and other workloads developed in a cloud into an on-prem environment, and 43% had reversed workload migration, moving them back to a physical data center from the cloud.
While nearly every organization said they rely on cloud services as part of their data protection strategy, one-third do not backup cloud-hosted file shares and 15% neglect their cloud-hosted databases, according to the survey.
The cost of security is growing, too.
Enterprise security and risk management spending is forecast by Gartner to grow 11.3% in 2023, with cloud security as the largest spending segment.
The market for cloud access security and workload protection will grow more than 25% to reach $6.7 billion next year, according to Gartner.
The price of failure is also high. The average cost of a data breach reached a global high of $4.4 million this year, according to IBM. American companies paid an average of $9.4 million per incident.
Cloud-based services were responsible for 45% of breaches in the 12-month period ending in March 2022, according to IBM.
Faulty assumptions are to blame for some of the confusion. When companies deploy “as a Service” features, they may, incorrectly, believe that built-in “undo” functions and server resilience measures are sufficient, the Veeam report said.
For commonly used enterprise applications, such as Microsoft 365, the problem is less severe. Only 4% of respondents rely solely on the built-in recycle bin or undo capabilities of Microsoft 365 for backup, and more than 2 in 5 use Azure Storage to ensure recovery.