Dynamic Pricing: How Jelastic Beats Amazon Every Time
Frequently, Jelastic is compared with Amazon Elastic Compute Cloud, and both clouds have a lot in common. Jelastic and Amazon EC2 are designed to make cloud computing easier. Both web interfaces decrease any possible issues while creating or managing environments. You have a wide choice of software stacks and have full control over your instances, being able to set any required configurations.
Many features and other benefits (similar or totally different) are provided by both clouds. And, as a result, sometimes it can be difficult for users to make the right choice. They need to consider all of the pros and cons. In this article, we would like to highlight one very significant difference between Jelastic and Amazon EC2 – the pricing system.
Pricing Model Overview
The main idea of Amazon EC2 pricing, is that you use a predefined amount of resources within the instances you add. To get slightly more resources, it’s required to buy an instance twice as large which results in paying double the cost. Even if your application consumes less resources than included in one Amazon instance, you still have to pay for the entire amount. Moreover, if you want to just resize your current instance, without data migration you should use EBS which costs significantly more than non-EBS AMIs.
On the opposite end of the scale, Jelastic provides very granular resources allocation. You can change the size of instances without data migration and extra configurations. Moreover, with Jelastic automatic vertical scaling, all unused resources are not charged at all.
Besides that, you get a versatile billing model, allowing you to choose between usage based pricing, fixed pricing, or a combination of both options (Hybrid Pricing). Subsequently, you pay only for the actual resource consumption.
Jelastic also provides automatic volume discounts, so as you use more resources, your pricing automatically decreases. Jelastic’s pricing is flexible, dynamic and automatically selects the correct volume discount as you use more resources. Jelastic ALWAYS provides the lowest pricing model for your application load – automatically.
Automatic Scaling for Price Optimization
Both Jelastic and Amazon EC2, provide an excellent feature, enabling resource scaling for pricing optimization, making it more user-friendly. However, this functionality is implemented in a diversely different way in our clouds, that also influences the scaling impact on the price. Amazon supports automatic horizontal scaling and Jelastic provides both – horizontal and vertical scaling.
When your application hosted on Amazon is growing, you need to upgrade to a larger instance in order to get additional resources – you can’t simply scale them vertically. Also, instead of starting a new instance and having to configure it from scratch, you can try to resize the existing instance by moving it to more suitable hardware. To achieve this, you are required to use EBS which brings more technical difficulties and much more expenses. In addition, you will face some outages while moving and your private and initial public IP addresses will be changed. Migration from a small instance to a bigger one also requires a reload of the virtual machine, which can cause data loss.
For production applications, you will most likely decide to use horizontal scaling. Amazon EC2 allows you to scale your capacity up or down automatically, based on the conditions you define. You need to state how many instances to add if the resource consumption raises. Sometimes, you have to reserve even more than you need, just because you can’t predict for certain, the needs of your application.
In its turn, Jelastic provides automatic vertical scaling of CPU and RAM (cloudlet = 128MB RAM and 200MHz CPU). You simply decide the maximum limit you are ready to consume and Jelastic automatically defines the optimal amount of resources required for your app.
With Jelastic, you get more superior granularity while scaling. In contrast to Amazon, Jelastic’s resource addition is performed automatically with small units (cloudlets). If you need to exceed the fixed scaling limits, this is performed manually with no need to reload the virtual machine or move to another server (which is usually required with Amazon).
You can also reserve a specified amount of Jelastic resources, if you are sure that your application will consume them. Such reserved resources are charged at a much lower price. By using this method, you don’t overuse unnecessary resources and as a result, you save your money.
In addition, you also can easily scale instances horizontally by adding additional servers. Horizontal scaling can be applied manually via the dashboard or automatically using Jelastic’s API. Soon, this process will be even more flexible, easier to use and automated – you’ll be able to set triggers for initialization of automatic horizontal scaling.
Cost Comparison for Real Applications
Now, let’s translate this theory into practice. Below we are going to present the numbers of actual resource consumption by the real applications and compare the expenses for those resources on both – Jelastic and Amazon EC2. Consequently, we will clearly show which cloud is more cost effective and suitable for your application, in terms of price.
As an example, we review four applications hosted in Jelastic in different parts of the world and covering different business sectors:
To maintain a fair comparison, we use Amazon prices in the same regions.
It is also interesting to note, that hardware settings at hosting providers of Jelastic correspond at least to m1.* settings on Amazon (this depends on the hosting provider). In addition, they are configured for fast I/O performance and compute optimized, due to use of SAN arrays with high-performance SSD and SAS storages that correspond to m3., c3. or i2.* amazon instances.
Hosting in South America
Our first two applications (Travel Service and Game Studio) are hosted in South America (Locaweb). Both of them are deployed to the environments with an application server and database.
For normal performance of the Travel Service application on Amazon, we need to choose m1.medium type of EC2 instance for the database and m1.large for the app server. These are 3.7 GB and 7.5 GB respectively. Subsequently, every year it will cost about $1,025 and $2,549 for each virtual machine.
While using Jelastic, if you only reserve the same amount of resources and do not consider any discounts, the price will be almost the same as for Amazon hosting.
But let’s consider all the benefits of Jelastic pricing!
Jelastic’s automatical vertical scaling is a unique technology that provides significant cost savings as you pay only for what you use. You don’t need to reserve the resources (more than you really need). Just set the maximum limit and Jelastic will consume only the required amount of resources.
The only reason to reserve resources in Jelastic, is to access more beneficial pricing, as our hosting providers charge the reserved cloudlets at a reduced price. You can reserve a minimal amount of resources to save even more.
Considering all of the specifications, applications like Onde.ir are charged $444 for database and $621 for app servers per year.
Comparing the pricing on both clouds, we can show that the annual cost savings with Jelastic is 52%
Jelastic hosting service providers (including Locaweb in South America) also provide discounts based on the amount of consumed resources – the more you use, the cost reduces. The benefits of this pricing model can be seen in the example of our next application – murastudio.com
Using Amazon, this application requires m1.small instance (1.7GB) for an application server and m1.medium (3.7GB) for the database. Thus, it will consume $1,533 from your budget every year.
At the same time, Jelastic hosting service providers, (in this case Locaweb) provides a flexible scale of discounts which help you to decrease your spending. With the abovementioned amount consumed in Jelastic, you will pay about $1,153 per year.
That’s approximately 25% saved using Jelastic.
Hosting in Europe
To compare the expenses of application hosting in Europe, we look at the Financial Fund application hosted at reg.ru (Russia). This app is deployed to the environment with four app servers.
To cover the needs of this app in Amazon, we take a m1.large virtual machine for app servers (7.5 GB each). For these resources, you’ll be charged $6,676 every year.
At the same time, with Jelastic you will pay just $3,283. This is the result of automatic vertical scaling and usage-based discounts implemented in our system.
It is notable, that even though the hourly price in Jelastic is higher, the total amount payable is 50% cheaper annually, than Amazon.
In the case that you use 50% of the resources, you pay only for them – not a dollar more.
Hosting in the USA
Finally, let’s also compare Amazon and Jelastic hosting in the USA. The subject of this experiment will be the High School application hosted at Webhosting.net with one app server and Maven build node.
From Amazon’s side, we’ll need a m1.small instance (1.7GB) for the application server and t1.micro (0.615GB) for Maven. Amazon’s yearly price will be $632.
Note that we don’t choose t1.micro because it has very poor performance of the disk IO and cannot be used for production purposes. That is why we preferred more powerful and production-ready m1.small which is minimal in the range of suitable instances.
Meanwhile, using Jelastic, this application doesn’t even consume the minimal amount of resources required with Amazon. And, with the help of Jelastic automatical vertical scaling and usage-based pricing, the owners of this app are saving a tremendous amount of money. The annual price using Jelastic is $155.
Now that’s competitive pricing and a massive saving – this is almost 73% cash, saved from your budget every year!
In conclusion, let’s summarize the pricing of Amazon and Jelastic, and therefore how much you spend and save.
- Offers only instances with reserved consumption limits
- Each next instance type is usually 2 times bigger and 2 times more expensive than the previous one
- You always pay for the upper limits, no matter how much your application consumes
- You can’t simply add more resources to a single instance in case of load spikes: reserve bigger instances or add more instances and pay for that additional charge
- Application is scaled by units of resource equal to 200MHz of CPU and 128MB of RAM
- You can set a minimum number of resources to be reserved for your application to get a discount and also set maximum limits in case your app is going to demand them
- Your application is billed hourly, based on actual resource consumption because of automatic vertical scaling technology
- Higher volume usage is automatically cheaper due to the automatic discounts
- Granular vertical scaling solves issues with load spikes
We hope that our examples clarify the differences between Amazon and Jelastic pricing. It’s up to you to decide which platform and geographical location is most appropriate and beneficial for your application. You can test the Jelastic platform and pricing system during a free trial period – Sign up for a 14 day free trial!
Watch the video about Jelastic pricing to get even more details.