Welcome to the final chapter of our journey! Over the past blogs, we've broken down every aspect, looked at the subtleties, and investigated how two possible solutions—AWS Elastic Disaster Recovery (DRS) and Veeam—compare with our client's needs.
Can review the case study here.
Now, let’s cut through the noise and get to the heart of the matter—because the right choice isn’t just about features or costs. It’s about finding the best fit for the client’s unique needs. Let’s dive in!
AWS-Native vs Third-Party Service
Now this is an easy one – ain’t it?
We’ve explored all the factors and compared two potential solutions: AWS’s DRS and the third-party Veeam. From our comparison table, it might seem like Veeam is the obvious winner. But is it really? While both solutions have their strengths and weaknesses across different factors, we’re fortunate that the client didn’t specifically request an AWS solution—just a workable one. With a few adjustments, either solution could meet their requirements. But in the end, we need to present one clear option to the client. So, which one will it be?
Before we pick, let’s explore one more factor—cost.
While not always explicitly stated as a requirement, the cost of the solution is a crucial factor to consider when designing a solution. After all, this will be a business expense for the client, so it’s important to provide not only a functional solution but also a cost-effective one.
When comparing DRS and Veeam, the cost structure for each is different. With DRS, there is a flat per-hour fee for each server being replicated to AWS, and you also pay for the replication instance(s), underlying volumes, and any recovery instances created during the recovery process in AWS. For each recovery instance, you incur charges for compute, memory, and storage.
On the other hand, with Veeam, you pay for the Veeam license for your on-premises environment. In addition to that, you pay for storage in S3 and API calls to and from S3. You’ll also incur costs for any recovery instances provisioned from the backups stored in S3. In both cases, recovery costs are only incurred when recovery is initiated into the AWS environment.
So back to our initial question – which of the two options are we recommending to our client?
Let’s go back to our comparison chart and see what the score is:
On the surface, it may seem like Veeam is the obvious winner. However, let’s go back to our case study and check if Veeam meets all our requirements:
Comprehensive Backups – _Yes _
Enhanced Recovery Capabilities – Yes
An RTO of 1-2 hours for the ERP System- No
Recovery options that include both on-premises restoration and the possibility of running the ERP in the cloud – Yes
Flexible RTO and RPO for Other Systems: Yes
The verdict? The ERP system is the outlier—and as the most critical system, we can’t ignore the need for a 2-hour RTO.
So where does that leave us? My vote? A hybrid approach. DRS for the mission-critical ERP system and Veeam for the more flexible Information and Library systems.
Of course, there’s more to this decision than meets the eye. A hybrid solution can bring added complexity and cost, so as an architect, your job is to present all viable options along with their pros and cons. In this case, we’re looking at:
- Veeam only
- DRS only
- Hybrid approach with Veeam and DRS
So, what do you think? Which solution would you have recommended to the client and why?
Top comments (0)