AWS is known for having very measured prices, but some services like RDS only charge you for the EC2 instance that the service uses for computer use. But these services charge up and you get less performance per dollar compared to a DIY solution.
The server is the service
This pricing model makes sense from the end of AWS. Let̵7;s take RDS as an example – a great service that makes it easy to run and manage databases. If AWS wanted to sell this as a traditional PaaS application, they would have a hard time pricing it as a single monthly payment.
Instead, AWS marks the price of the servers that the service runs on top of and places them in special class that is specific to the service. For example,
db.m5.large. This selection can be one pulp. For RDS, the price is one
m5.large the instance goes up from $ 0.096 per hour to $ 0.171 per hour – an increase of 80%. For memory optimized
r5 the range, the essay is 90%.
When you pay twice as much per month to use a managed service, it may make more sense to do it yourself on EC2. Databases are not that difficult to set up, and even if the burden falls on you to keep up with maintenance, the price reduction means that your dollar will move towards more performance if you do it yourself. You can run two
r5.2xlarge case for the price of one
Since all this is said, there is still reason to use AWS’s managed services. If you are a large company, running databases is a huge task that involves a lot of staff doing things that AWS RDS can do for you. You have to pay that staff, so when labor costs outweigh the price increase from core to core, RDS gets cheaper at a certain breaking point.
However, if you are the owner of a small business, you should probably consider doing it yourself and not play for the expensive business solution.
How much extra will you pay?
Here we examine how much marking AWS adds to each service. Prices are calculated based on current
us-east-1 (N. Virginia) Pricing.
RDS—AWS Premium Database Service. Public instances are marked with 80%, and memory-optimized instances are marked with 90%. If you care for reserved instances, even on a three-year contract, the RDS is still up 55%.
DocumentDB-AWS’s clone of MongoDB, is offered as a managed service. One
r5.2xlarge instance costs 120% more than EC2. And EC2
r5.4xlarge instance costs less than DocumentDBs
Neptune-AWSs premiumgrafidatabas. It is 176% more expensive than EC2. It’s worth noting that Neptune is proprietary, so you can not run it on EC2. But open source databases like Neo4j are available as an option.
aurora,Aurora is already more expensive than RDS and is particularly bad and starts 130% more expensive than EC2 on demand. A three-year period helps a bit, with the cheapest you can drive an Aurora at 87%. There is “Aurora Serverless”, but it’s really just an autoscaling configuration for Aurora. AWS really offers a serverless database in the form of DynamoDB, but it is difficult to say how much more expensive Dynamo is compared to EC2, as it charges based on usage alone.
Fargate-AWS’s service for running managed containerized applications without worrying about servers. You simply pay $ 0.04048 per vCPU and $ 0.004445 per GB per hour. This makes direct comparisons a bit difficult, but adding the specifications for the general purpose 16 core 64 GB
m5.4xlarge instance amounts to $ 0.93216 per hour. Compared with
m5.4xlargeis the market price of $ 0.768 per hour, that is an increase of 20%. However, Fargate applications only use as much resources as they need, so you will probably reduce the margin a bit. Overall, Fargate is priced pretty well for what it’s worth, and you’ll see even greater savings with Fargate Spot Instances.
Elasticsearch-A service for searching log files. It runs on EC2 at a 47% mark. Despite the resounding name AWS-y, Elasticsearch is open source and developed by Elastic NV, so you can run it yourself on EC2 for much cheaper.
SageMaker-AWS’s service for running machine learning in the cloud. You typically use GPU-accelerated instances like AWS
p3 line up. Pricing is divided into a few sections – for construction models it is an increase of 40% compared to EC2. The education models are the same increase by 40%. However, in both of these scenarios, you will only be charged when the instance is actually used to build or train models, and not for all installation time associated with running it yourself. With the high hourly costs for
p3 lineup, this compensates the price enough to make SageMaker cheaper in practice. For at least training models, when you run your model in production, you see the same increase of 40% to do everything in SageMaker.
Elastic inference –Running ML inference is often much cheaper than running training, so you may not need a full Tesla V100 GPU to deploy your ML models. AWS offers GPU accelerators that can be connected to running EC2 instances. The cheapest
p3 GPU, den
p3.2xlarge, costs $ 3.06 per hour to run and comes with 8 cores, 61 GB of RAM and 16 TFLOPS GPU performance. For an accurate comparison with GPU-only EI accelerators, subtract the vCPU and RAM costs. It similarly specified
m5.4xlarge costs $ 0.768, so the estimated cost of a single V100 GPU is about $ 2,292, give or take, which is about $ 0.143 per TFLOP. The cheapest EI accelerator, which provides a single TFLOP performance, costs $ 0.120 – 16% reduce above the EC2 price. Option 4 TFLOP is even better: a 40% reduction compared to EC2. If you run machine learning, EI is the way to go.
Light Sail—AWS for beginners. Lightsail is designed to have pricing without nonsense and offers simple monthly plans and a much easier way to manage servers for the less cloud-savvy people out there. The cheapest plan comes at just $ 3.50 per month for 512 MB of RAM, 2 vCPUs and 20 GB of SSD space. ONE
t2.nano instance would cost $ 4 per month, with the 20 GB disk costing $ 2 per month, making Lightsail a 41% saving at this level.
Going up to something more reasonable gives different results. The $ 80 level includes 8 vCPUs, 16 GB of RAM and 320 GB of SSD space. It similarly specified
a1.2xlarge instance costs $ 153 to run, and the EBS volume costs $ 32. Another win for Lightsail with 56% cost savings. Sure, you can probably see this type of savings with either reserved instances or location instances with automatic scaling, but Lightsail requires no commitment or technical installation.