When I was a kid my mother told me never talk about politics, religion or money. Luckily, she'll never read this. I'm pretty sure she thinks I sit in a basement and do something that involves backing up iPhones. I can promise we won't touch politics or religion though, but we are most certainly going to talk money, and more importantly, how to help reduce IT cloud spend.
All three major cloud vendors (AWS, Azure, and GCP) offer native backup options. The main advantage being ease of deployment as it's just a checkbox to enable for the most part. As companies move workloads or create net new workloads in the cloud it's common the data protection team is left behind as cloud admins will manage the backups themselves. This quickly becomes unsustainable though both from a cost and management perspective.
Sifting through monthly cloud bills to find backup costs is not as easy as one might think. To understand the total cost of ownership (TCO) associated with protecting cloud workloads, management needs to know what cloud vendors charge to protect the workload and storage the backups consumes on the backend. Finding that backend storage cost is not always so simple. To help give you a snippet below is the monthly cost to protect workloads using the native backup tools cloud vendors offer.
This can quickly become very expensive. For example, protecting an 800 GB SAP machine in Azure would cost nearly $2,500 a year! Protecting a 1 TB SQL machine in GCP would cost about $1,000 a year. Now, AWS looks much more affordable at first glance but services like EC2 cannot be backed up to S3 with AWS Backup, so the storage costs become hefty. Not to mention the problem of snapshot sprawl as many of these snapshots are forgotten about and linger for months or even years longer than intended.
Let's look at a TCO in a larger scale exercise to help paint a better picture. Let's assume:
-
We have 130 cloud VMs that need protection
-
Each workload is 1 TB
-
Has 5% change rate
-
30 day retention
-
Backups do not get dedupe and compression with native tools as stated on their sites
Azure Backup TCO Example:
As we can see below the total yearly cost for just 130 workloads quickly adds up with native Azure backup. For the nerds (like myself) who want to know how I came to these numbers, I added up the full backup of each workload plus the total for incremental backups for storage costs. In the Azure VM row that would be 250 TBs *.0224 = $67,200. Then I added the backup costs to protect 100 Azure VMs which is 100 instances * $20 *12 = $24,000 which is a total of $91,200. Remember Azure charges by the 500 GB increments, so the backup charges are $20 per month for a 1 TB machine.
GCP Backup TCO Example:
GCP is around the same cost at Azure. I won't bore folks with the basic math equations again. Think you get the gist.
AWS Backup TCO Example:
AWS is much cheaper in comparison as there is no backup costs. Protecting workloads in AWS is free, but the storage is nearly 2.5x the price!
Veeam Backup TCO Example:
The goal of this post isn't to scare folks from ever protecting workloads in the cloud. It is to show a cost effective solution while not sacrificing any simplicity or functionality. If anything also gaining functionality.
The cost savings by using Veeam to protect cloud workloads can be huge for several key reasons:
-
Storage costs are minimized with a reliable 2:1 dedupe/compression
-
Costs associated with snapshots to backup workloads can be significantly reduced if not eliminated altogether if Veeam's native cloud agents are used
-
It is the same reasonable backup cost to protect a cloud workload regardless what application it is running or how large the instance is
There's more to life than money though. As alluded to earlier, there are also additional capabilities Veeam provides that native cloud backup does not. The major differentiators I see resonate are VPC backup, data mobility cross cloud platforms (restoring an AWS EC2 to an Azure VM and vice versa) and ease of management under a single UI. Below are additional capabilities for each cloud platform.
It's never fun to talk about money, but in this economy it's important to find ways to save yourself and your company money. Protecting cloud workloads with Veeam is the best of both worlds - saving money without sacrificing functionality. Oh and please don't tell my mother about this post.
