One of, if not the most, challenging aspects of cloud cost optimization and data is achieving your desired outcome without being shocked by your bill at the end of the month. Balancing the needs of your business and your proverbial checkbook is not an easy thing. In fact, an entire practice called FinOps — a portmanteau of finance and DevOps — was born just to address this, bringing together business and technical teams to collaborate on cloud cost optimization and efficiency.
What is Cloud Backup Cost Optimization?
One area where this is of particular importance is cloud backup. We’re dealing with copies of large data sets that are often stored for lengthy periods of time. Flexera’s 2023 State of the Cloud Report demonstrates this perfectly, with businesses ranging in size from SMB to enterprise ranking cloud spend management as their top challenge.
Understanding Cloud Backup Costs
To achieve cloud cost savings, it is critical to better understand the transition from typical CAPEX models to the pay-for-what-you-use OPEX models of the cloud. Cloud backup consumes varying degrees of compute, storage, and data transfer resources, not to mention the licensing involved with the backup tool itself. Sometimes vendors will combine all the above into one, irrespective of how many resources you consume, representing savings for some and additional costs for others. Other vendors are more transparent, with licensing fees only entitling licensing, and with storage, compute and data transfers billed separately based on their exact usage.
So, when it comes to cloud backup, what are some best practices to meet required service-level agreements (SLAs) like RTOs, RPOs and retention without breaking the bank?
Strategies for Cloud Backup Cost Optimization
So, what are some best practices when it comes to cloud backup to meet required service level agreements (SLAs) like RTOs, RPOs and retention while realizing cloud savings, whether you’re an end user or service provider?
Cost-Effective Cloud Storage Options
Perhaps the most obvious resource consumption for cloud backup is storage. But not all storage options are created equal, with varying performance and cost needing to be balanced based on the criticality of data, how long it needs to be retained, and how old the data is. The visual representation below gives a general comparison of AWS list prices for common storage types used in backup based on pay-as-you-go, per-GB, per-month rates in us-east-1, accurate as of Aug. 22, 2023. Azure and Google follow a very similar model, and understanding the pros and cons of these storage classes is critical to cloud cost optimization.
Snapshots vs. Backups
Native snapshots are typically the go-to method for protecting cloud-hosted resources, whether they are manually created and managed, scripted, or orchestrated with another tool. Snapshots afford us the ability to achieve low RPOs as they can be taken frequently, as well as low RTOs to overcome data loss scenarios quickly. However, from a cost lens, you can see above that the volumes snapshots are stored on are among the most expensive cloud storage options. A true backup can be made from these snapshots, compressed, and then stored on more cost-effective object storage to meet compliance, retention and budgetary requirements.
Leverage Cost Calculators
There are a lot of resources out there to help you understand cloud costs based on your business needs, such as the AWS Pricing Calculator and Azure Pricing Calculator. Cloud providers are incredibly transparent about their pricing and often strive to help you drive the best value for your business. One area where this can become a little muddy, however, is with backup. Veeam’s solutions for AWS, Azure and Google Cloud all come with proactive backup cost calculators built directly into the policy wizard to help you get an understanding of what that policy is likely to cost you monthly, which helps avoid bill shock. It also breaks down the total cost into subcomponents like snapshots, backups, replicas, archives, traffic and API calls, so if tweaks do need to be made, you’re already in the right place to make those changes.
Maximize Discount Opportunities
Much like anything you buy these days, the more you purchase, the better the price per unit. This holds true for not only the data protection software used for protecting cloud, but the cloud resources themselves. AWS, Azure and Google all offer commit programs to a predetermined annual expenditure that results in contractual discounts on consumed resources, similar to an enterprise agreement. As you migrate to and grow within the cloud, forecast your spend routinely to not only avoid bill shock, but to negotiate volume-based discounts early.
Also worth noting here are the differing types of compute instances not just from a power perspective, but also pricing structure. Workloads with consistent utilization typically benefit from reserved instance pricing. Those with consistent utilization but with predictable times of increased performance requirements benefit from auto scaling, and workloads that are dynamic and short-lived typically benefit from on-demand instances.
Data Lifecycle Management
Efficient lifecycle management of your backup data ensures that backups will remain accessible and usable when needed while minimizing the financial impact associated with maintaining storage for unused, infrequently accessed, or unnecessary snapshots or backups.
By implementing a well-structured approach, organizations can optimize storage resource utilization while reducing cloud costs, typically by transitioning backups to less expensive storage tiers as they age, where the likelihood of a recovery diminishes, but data must be retained for compliance purposes. Here’s a cost comparison of a 100GB Amazon EC2 instance with a two-year retention requirement between snapshots alone and a tiered approach where data is tiered to Amazon S3 after a month, S3 Glacier after six months and S3 Glacier Deep Archive after a year. Data stored on Amazon S3 storage classes are image-based backups with 2X compression.
Deduplication and Compression
One of the many benefits that using a true, image-based backup to complement a snapshot is the ability to deduplicate and/or compress that backup data. Deduplication removes redundant data blocks within a backup file, whereas compression removes and reduces the bits required to represent the data within the backup. As mentioned above, Veeam is capable of 2X compression for backups of cloud-hosted workloads, reducing the required capacity by half, which in turn has a significant impact on the associated spend.
Automated Backup Policies
Automation plays a critical role in reducing cloud costs for several reasons, as well as aiding governance, compliance, alerting and reporting. Firstly, cloud backup automation built on policies ensures the right degree of protection and retention is applied to the right workload. For example, dev workloads seldom require the same rigor of production, mission-critical apps. A one-size-fits-all approach would either result in over protection and overspend, or under budget at the cost of inadequate protection. Implementing several policies with varying backup frequencies, data lifecycle management automations and retention durations that are then applied to varying criticalities of workloads based on tags helps ensure adequate protection on budget. See this example below.
There’s also a human element to automation. Streamlining data protection processes eliminates the need for dedicated personnel to manage manual backups, freeing up time and resources for them to focus on other strategic priorities while still getting the protection job done.
Monitoring and Reporting Tools
Once you’ve taken control of your costs, you need to keep control of them. This is where Day 2 operation feature sets come into play. You’re now over the hurdle of deployment and initial implementation — but this is no time to take your eye off the financial ball, since cloud creep is real. Remain focused on not just maintaining, but continually optimizing operations as your environment evolves and grows. Monitoring toolsets are critical, but the process isn’t always straightforward, especially in the hybrid-/multi-cloud world we live in. Be sure to use solutions that can span the breadth of all your environments and make your life easier by centralizing management, monitoring, and governance into a single location.
Real-Life Case Studies
Veeam Data Platform is the industry-leading Modern Data Protection solution to help you protect, secure and manage all your data across your hybrid-/multi-cloud environment. Whether you’re sending backups of on-premises workloads to the cloud or protecting cloud-hosted IaaS, PaaS and SaaS data, Veeam can help you do it all easily from a single platform. Comprehensive cost optimization capabilities that are built-in throughout can help you take and keep control of your cloud expenses without sacrificing your degree of protection or how quickly you can recover.
You can learn more about how HBC saved over $1 million in CAPEX and $600,000 in OPEX annually by implementing Veeam across its multi-cloud implementation of on-premises, AWS, and Azure. Or, how CIOX Health manages a monthly spend of $20,000 for all backup resources when protecting 500TB of patient data daily from their colo to AWS, AWS to Azure, and Azure to AWS.
Controlling and optimizing spend in the cloud has proven to be the leading challenge for organizations large and small in 2023, especially in today’s economic climate. This is especially true for data protection and the necessary copies of data sets that need to be made and stored for lengthy periods of time. However, getting on top of your cloud spend is not a lost cause, as there are plenty of best practices and toolsets from first and third parties available to provide you with the necessary insight and technology to achieve all your outcomes. While it may seem daunting at first, your future self (and your boss) will thank you later for taking the first step today. After all, who wants to be the person who can’t keep their business running because they were penny wise and pound foolish?