Conducted by Zerto, a provider of backup and recovery tools, and Enterprise Strategy Group (ESG), the survey found three-quarters of respondents (75%) expect container-based applications to be backed up in the same manner as every other application. The average recovery time objective (RTO) tolerance cited by survey respondents today is 2.87 hours, while the tolerance for a recovery point objective (RPO) is on average of 22 minutes.
Deepak Verma, director of product strategy for Zerto, says that it should be apparent by now that existing RTO and RPO goals are not going to meet the requirements of container applications that need to be able to recover from a disruption in seconds and minutes. Zerto is currently making available in beta a backup and recovery platform that runs natively on Kubernetes clusters to reduce backup and recovery time for container applications.
Backup and recovery of container applications is becoming a bigger issue as the number of stateful applications based on containers continues to steadily increase. Backing up and recovery for containerized applications will also be more challenging because they will be running on multiple distributions of Kubernetes running on multiple clouds, notes Verma.
Nearly three-quarters of survey respondents (71%) said they have deployed or plan to deploy container-based applications as part of a hybrid cloud strategy, which Verma says is why 57% identified multi-cloud support as one of the most important features for backing up container environments. Many survey respondents are already aware that managing backup and disaster recovery in either a container environment spanning hybrid clouds (44%) and/or across multiple public clouds (39%) represents a major challenge.
Nevertheless, less than a quarter of respondents (21%) said they have or plan to deploy container applications only on a public cloud.
Given those challenges, Verma says it’s apparent backup and recovery will need to become a lot more automated within the context of a DevOps workflow. At the moment, however, it’s not clear how much DevOps teams are moving to extend their workflows to include backup and recovery.
Most backup and recovery tasks today are handled by IT administrators using graphical tools. With the rise of ransomware, it’s become more critical than ever for IT teams to have pristine copies of data readily available. The challenge is that much of the data backed up isn’t fully recoverable for one reason or another. Worse, yet, that data may already have been encrypted by ransomware before it was backed up. In theory, embracing DevOps processes to automate backup and recovery would also make it a lot easier to test whether backups had been made successfully.
With the arrival of each new platform, data protection issues eventually surface. There’s no shortage of ways to protect container applications. The real challenge is remembering to make it a priority before it’s too late to do anything meaningful about it.