Reducing Your Database Hosting Costs: DigitalOcean vs. AWS vs. Azure
Monday, May 11, 2020 at 11:09PM
If you’re hosting your databases in the cloud, choosing the right cloud service provider is a significant decision to make for your long-term hosting costs. This is especially apparent in today's world where organizations are doing whatever they can to optimize and reduce their costs. Over the last few weeks, we have been inundated with requests from SMB customers looking to improve the ROI on their database hosting. In this article, we are going to compare three of the most popular cloud providers, AWS vs. Azure vs. DigitalOcean for their database hosting costs for MongoDB® database to help you decide which cloud is best for your business.
Comparing Cloud Instance Costs
So, which cloud provider provides the most cost-effective solution for database hosting? We compare AWS vs. Azure vs. DigitalOcean using the below instance types:
| AWS | EC2 instances |
| Azure | VM instances |
| DigitalOcean | Droplets |
Since database hosting is more dependent on memory (RAM) than storage, we are going to compare various instance sizes ranging from just 1GB of RAM up to 64GB of RAM so you can see how costs vary across different application workloads.
Let’s take a look at the monthly cost (720 hours) of database hosting for standalone, on-demand, dedicated instances on AWS, Azure and DigitalOcean. As you can see from the graph below, DigitalOcean database hosting provides significant cost-savings over both AWS and Azure. Additionally, their Droplet pricing is extremely simple and easy to understand - $5/GB.
As you can see from the above chart, on average, DigitalOcean instance costs are over 28% less expensive than AWS and over 26% less than Azure.
Comparing ScaleGrid Database Hosting Costs: AWS vs. Azure vs. DigitalOcean
As mentioned above, the reason we decided to write this article is because of a recent increase in questions from customers on how they can reduce their database hosting costs, so we wanted to make sure to compare the costs of our fully managed DBaaS solution across cloud providers as well. Here are the configurations for this comparison:
| Plan | Dedicated Hosting |
| Database | MongoDB® Database |
| Replication Strategy | 2 Replicas + Arbiter |
Our Dedicated Hosting plans are all-inclusive, including all machine, disk, and network costs, as well as 24/7 support. These plans are fully managed for you across any of these cloud providers, and comes with a comprehensive console to automate all of your database management, monitoring and maintenance tasks in the cloud.
Let’s take a look at how ScaleGrid Dedicated Hosting pricing compares across AWS vs. Azure vs. DigitalOcean:
| ScaleGrid Dedicated Plans | AWS | Azure | DigitalOcean |
|---|---|---|---|
| 2GB | $190 | $187 | $104 |
| 4GB | $330 | $374 | $140 |
| 8GB | $657 | $750 | $300 |
| 16GB | $1,164 | $1,250 | $500 |
| 32GB | $1,912 | $2,025 | $800 |
How much can you save migrating to DigitalOcean?
So, are you’re deploying MongoDB® database on AWS or Azure, and wondering how you can lower your database hosting costs? Let’s see how much you can save by migrating your hosting for MongoDB® database to DigitalOcean:
ScaleGrid’s Dedicated Hosting service with 2 Replicas + Arbiter for MongoDB® database on DigitalOcean saves you on average 122% on your monthly AWS hosting costs, and 140% on your monthly Azure hosting costs. The above chart outlines the cost savings across different plans, and ranges from around 80% cost-savings for 2GB of RAM, up to 153% cost-savings in our 32GB of RAM plan size.
DigitalOcean Advantages
DigitalOcean provides many advantages for database hosting, and you can learn more about them in our The Best Way to Host MongoDB on DigitalOcean blog post. Here’s a quick overview of the key advantages:
- Developer-friendly
- Simple pricing
- SSD-based VMs
- High performance
DigitalOcean Hosting FAQs
Is my database cluster still highly available?
Yes. All of our high availability options are offered in DigitalOcean, including 2 Replicas + 1 Arbiter, 3 Replicas and custom replica set setups. DigitalOcean does not have the concept of availability zones (AZ), so we distribute the nodes across different regions. For example, in the US, we distribute nodes across New York 3, New York 2 and New York 1.
Does it affect latency?
Yes, you can see an increase in latency. Ideally, we would want to see both the application and the database in the same datacenter. So, if you're hosting your application in AWS or Azure and move your database to DigitalOcean, you will see an increase in latency. However, the average latencies between AWS US-East and the DigitalOcean New York datacenter locations are typically only 17.4 ms round trip time.
How can I migrate?
ScaleGrid provides an Import wizard to migrate data from one cluster to another. If you have any special needs for your migration, please contact support@scalegrid.io.
|
Azure, Cloud Provider, Dedicated Hosting, DigitalOcean, Droplets, Hosting Costs, MongoDB, ROI, VM Instances, aws, database, ec2 in AWS, Azure, Database, DevOps, DigitalOcean, MongoDB, best platform, clusters, cost savings, deployment, hosting, public cloud Tweet
Reader Comments (10)
This is quite a sad marketing post.
I am expecting better from this blog, not shameless marketing plugs.
It's just a marketing post. I don't see any good points on it.
The graph shows 150% savings. Maths just doesn't work that way. If something was $100 and I bought it for $50 that is a 50% savings - not 100%.
150% savings implies I got it for LESS than free.
Agree with Diego! Instance cost is only a portion of running a production system.
If you split your system across multiple providers, you also need to think about
* monitoring - does your alerting and monitoring provide insight into both platforms? How well can you correlate events across platforms?
* tooling - monitoring tooling, orchestration tooling, IR tooling
* processes and training - are your team members trained on both platfoms?
* procurement and vendor management - (you definitely do this right?) Now you have two vendors.
* network - you mentioned latency, but what about data transfer costs? Routing? Network integration (because you're not exposing your DB to the internet are you?)
Indeed, a disclaimer at the top that this is a sponsored post is the least I expected.
That's because jw it's not a sponsored post. It's just a post someone added to the site, just like anyone can add at any time. You can add a post if you want. I don't let a lot through because you wouldn't believe what people post, but these are on the edge so I let them through. There's not a lot else going on.
If a post is about a vendor-specific topic, then fine, but the title should clearly reflect that. A more apt title in this case would be "Reducing MongoDB Costs Across Cloud Providers with ScaleGrid." A deliberately misleading post like this not only wastes everyone's time, but it also damages the reputation of both this website and the author's brand. So, for the benefit of everyone involved, I would suggest enforcing a policy where vendor-specific posts require the vendor to be clearly identified in the title.
this blog is going downhill fast ... there's less and less decent content. it's hardly worth coming back for.
Life is complicated cowper.
Good that DO got compared with AZ, AW, but where is GCP in all of this?