Welcome to part two of our blog series on VMware data protection best practices. In part one, we talked about getting back to the basics for backup and what you can do to make protecting VMware easier. Now that you’re familiar with the basics, this blog will help your organization plan for the future. We’ve seen a big push towards cloud and VMware’s announcements about VMware Cloud. But what are the best practices for someone looking to move on-premises VMware to the cloud? Below, we’ll lay out this process in three simple steps.
Step 1: Simplify backup of on-premises VMs with the cloud
Moving to the cloud doesn’t necessarily mean moving all of your production workloads. In fact, the best approach is to start migrating non-critical workloads. Some start by moving their dev/test workloads to the cloud as they can quickly spin up and down these environments, reducing cost for idle resources.
But you don’t have to go to the cloud to see workloads being spun up and down, or waste resources. Even with on-premises VMware, it is easy to quickly spin up new applications within VMs, but the backup infrastructure is less dynamic. As new VMs are brought online, how should you ensure your backup infrastructure can protect them? You either need to scramble to expand your backup infrastructure or waste money on unused infrastructure. When you have multiple sites, how should you centrally manage all backups and the infrastructure at those locations? Before moving production workloads to the cloud, moving backups to the cloud can solve problems today and lay the groundwork for the future.
There are different approaches to moving backups to the cloud. One of the most common is to lift-and-shift your entire backup infrastructure to the cloud and run it within your own account. By doing so, the process will look and feel exactly as it did on-premises, albeit with a few potential complications. The first is that running in the cloud can be expensive, especially accounting for compute and storage costs. Even with cheaper object storage, things like egress costs to restore your data from the cloud and request pricing (not just the monthly cost of storing the bits) become the responsibility of the customer. And finally, the cloud has a very different security model than on-premises, and it takes time to become proficient — the last thing you want is a data breach of your backups!
Instead of lift-and-shift, if you really want to simplify your management and reduce cost, you should consider a cloud-native backup solution. This will allow you to stop worrying about the backup infrastructure, pay for what you store, and some offer transparent pricing, so you won’t be surprised by any costs (like doing restores). This approach simplifies managing VM backups and gives you centralized visibility across your entire environment, all without any hardware, while allowing you to get comfortable with the cloud.
Step 2: Reduce CapEx by eliminating your disaster recovery site
It is important to have a disaster recovery (DR) site in case something happens to your production data center. In the event of an emergency, you want to ensure that your workloads can recover and keep your business running. While everyone would like their application replicated in real-time, often only mission-critical applications are protected. But until the disaster strikes, your DR site sits idle, consumes power, and its duplicate infrastructure incurs additional expenses. Ideally, you want to periodically test that your workloads can recover in your DR location, but it isn’t always easy to spin up an isolated test environment. The last thing you want when a disaster happens is to realize you forgot to replicate part of an application and now nothing works! Once again, this is where cloud can help you.
Do you know where the largest “data center” with on-demand capacity exists? The public cloud! Cloud is perfect for scenarios where your workload needs to be spun up only when a disaster occurs. No longer do you need to maintain infrastructure that sits idle — helping you save cost. You can also periodically spin up and test your DR environment before a disaster occurs, gaining confidence that everything will work. But how do you get data up to the cloud and offer DR for more of your applications? By leveraging a converged backup and DR solution, you can provide everyday restore needs for your end-users, meet compliance needs, and provide a low-cost DR option by being able to failover backups into the cloud. Remember, you want to simplify management, so make sure the solution provides orchestration (i.e. runbooks). You can provide automation,be ready to go when disaster strikes, and also periodically test that everything will work. By doing so, you can leverage the cloud to eliminate the cost of maintaining a separate DR site.
Step 3: Unify and scale data protection for VMware Cloud hybrid
Now that we’re comfortable with some workloads running in the cloud, let’s start moving your production workloads to VMware Cloud on AWS. You can still use all the familiar VMware tools you know and love (vSphere, vCenter, VSAN, and NSX-T), but without the infrastructure management by VMware offering it as a service. You get the opportunity to focus on your day-to-day tasks — leaving the hardware maintenance and software patching to VMware. But now that you are moving some workloads to VMware Cloud on AWS, how do you protect it? You want a solution that scales with you — as you move more workloads to VMware Cloud on AWS, your backup solution scales. In addition, by moving from a CapEx to an OpEx model, you also want to pay for your backup solution in a similar fashion (refer to Step #1).
There is one key thing to remember about moving workloads to VMware Cloud on AWS — they don’t all have to go! You may keep some production workloads on-premises or burst others to VMware Cloud on AWS for a brief period before migrating back. VMware offers you the flexibility to keep your VMs either on-premises or in the cloud. The world is quickly transitioning to a hybrid approach, and when it comes to data protection, you need to ensure your backup solution can handle protecting your VMs no matter where it runs. You don’t want two siloed backup solutions, nor do you want to deal with managing a single solution in different manners across your environment (or managing backup infrastructure). You want a single solution that works seamlessly. If your VM moves from on-premises to VMware Cloud on AWS, your backup solution should automatically detect and protect it without you having to lift a finger. The best way is to have a cloud-centric solution that works across on-premises and the cloud to eliminate backup infrastructure management (think about the future!).
Cloud and VMware are not mutually exclusive — VMware is, and will continue to be important for many customers. While moving to the cloud can seem daunting, think about smaller steps to get you there. Start with less critical workloads like test/dev, backup, and DR to get comfortable with the cloud. As you gain more confidence and start moving your production workloads to the cloud and back, make sure your backup solution can work seamlessly to protect your VMs no matter where it runs. Druva offers a SaaS -based solution that helps you seamlessly protects VMware on-premises and VMware Cloud on AWS, and allows you to finally retire your DR site and start leveraging AWS for failover. Feel free to try out Druva and see what it can do for you. Don’t forget, you aren’t the only one on this journey!