Virtual data protection: Seeing is believing

Virtual data protection is hard, but could be easier if you had a better view into what’s going on. Learn how to increase visibility into your virtualization protection solution.

According to a recent Enterprise Strategy Group report on Trends in Protecting Highly Virtualized Environments, virtualization rates as one of the top (or top) challenges in data protection. But digging deeper, one finds that five of the top six challenges in protecting virtual environments are related to visibility. Specifically, the top six challenges are:

  1. Recoverability of data (disappointing but not surprising)
  2. Validating backup success
  3. Validating recovery success
  4. Identifying factors impacting performance of backup operations
  5. Simplified views across virtual infrastructure
  6. Response time in troubleshooting backup failures, policy violations and so on

Looking more closely at the root words — validating, identifying, views, troubleshooting — you can see that almost everyone is struggling with visibility. Consider this:

  • When you’re protecting a physical server, you can easily understand every component within the solution, from the installed backup agent on the physical server across an identifiable network path to a backup server, with defined software and storage.
  • When you’re protecting a virtual machine, abstraction layers are added at the hypervisor layer, as well as two tiers of “agent-like” functionality between the guest-based components alongside Microsoft Volume Shadow Copy Service (VSS), but also the interaction of the host APIs of vStorage/vStorage APIs for Data Protections or VSS and two operating systems, each of which could affect the functionality of the backupprocess.
  • When you’re protecting a private cloud, additional uncertainty is added because those virtual machines (VMs) are now dynamically created during elastic workloads based on overall load and usage; they can be provisioned through self-service portals and the VMs will move between hosts, clusters and datastores. Now, the backup software is truly trying to protect a moving target.

It shouldn’t surprise anyone that the top virtualized infrastructure protectionchallenges are visibility related. Too often, when a backup application is perceived as inadequate for virtual data protection, the presumption is that the backup software needs to be changed. But if we ignore the obvious differences in backup products (e.g., manageability, deduplication and so on), the reality is that most backup products leverage a hypervisor vendor’s approved methods for backing up VMs. Are there differences between products and methods? Absolutely. However, others would suggest you should discard your on-premises software and use a data protection service.

But if the challenges you’re experiencing are visibility related (or would be resolved with better visibility), then perhaps the answer is not to change software or services but to increase the visibility into your virtualization protection solution. Large enterprises discovered a long time ago that they often have to sustain multiple backup products for a variety of reasons, so they use one of a very small handful of monitoring solutions designed to scrutinize diverse data protection products and provide a consistent dashboard and view. Some of those tools not only provide data protection monitoring (DPM) across backup tools, but are optimized specifically for virtualized environments. When users were asked about the benefits of DPM tools, compliance, improved troubleshooting and assessing success/failure of jobs were the top recognized benefits.

Said another way, IT organizations that choose to add enhanced monitoring tools may actually start to realize the ROI of what they’ve already paid for and have installed because they can get their existing data protection product(s) to work better. Interestingly, this also provides some business opportunities:

  • Service providers (SPs) are often looking for new services to offer, and sometimes Backup as a Service (BaaS) is their choice. But SPs may not have the technical experience to deliver a solution that is more complicated than an all-off-premises Software as a Service or Infrastructure as a Service offering. Instead, BaaS has cloud components, but also requires on-premises agents, and recurring interaction between SP personnel and every subscriber’s IT team.
  • Local integrators (VARs) are often looking to evolve from transactional hardware-plus-software offerings to services with recurring revenue and more strategic value. But this may require urging the customer to rip out the solution that very VAR installed. Moreover, not every VAR is culturally equipped to stand up its own infrastructure (even in a collocation facility) and offer services.

Perhaps the answer for those SPs and VARs is to reconsider jumping into the BaaS market, and instead to enter the Monitoring (of Backups) as a Service market. This is particularly ideal for VARs who have the technical staff to remotely manage/monitor the very backup solution they installed earlier. The user doesn’t have to change their infrastructure, and their current investment provides the capabilities and value they originally thought they were purchasing.

Modern data protection is hard. Highly virtualized and private cloud protection is much harder. The next time you’re dissatisfied with the results of your current backup offering, first consider if you have the right visibility, including the correct processes and people, to ensure the product can perform as advertised. Then think about how much more strategic you could be if you didn’t have to babysit your backup solution because someone else, who is legitimately more experienced, is doing it for you.

 [Originally posted on TechTarget as a recurring columnist]

Leave a Reply