Estimate Your Cloud Data Transfer Costs: 3 Use Cases + Netflix

Oct 29 2015 | by Yoav Mor

0 Flares Twitter 0 Facebook 0 LinkedIn 0 Email -- 0 Flares ×

The cost of hybrid cloud management doesn’t end with resource costs, it includes the costs of transferring data into and out of the public cloud, as well. In addition, users that download large media rich data files such as movies also have a great impact on website owners’ public cloud costs.

In this article, we will use a few use cases to provide a framework of guidelines regarding which parameters should be taken into consideration when calculating data transfer costs into and out of a cloud environment.

Data Transfer Scenarios

There are multiple cases of moving data into and out of the public cloud, however, the three following scenarios reflect typical cases of data transfer.

#1 – Bursting
Bursting scenarios deal with an environment that is built entirely on-premises. The databases, applications and middle-tier servers are all based in the same location. Applications can be types of traditional stacks or modern web scale. The public cloud comes in handy when dealing with peak levels of demand by providing a “burst” of resources.

#2 – Disaster Recovery (DR)
DR refers to the replication of data and applications in the public cloud. Think about it like a spare wheel for your environment if anything goes wrong. The public cloud offers the option to create an environment on hold – “pilot light” configuration, meaning data is continuously backed up even though you are not paying for the environment to run.

#3 – Rich Media Sites
Lots of large media providers, from Netflix all the way to Glide, see the public cloud’s flexibility as a major advantage when it comes to storing large amounts of data and files. However these sites also need to deal with user downloads and streaming rich media files, which significantly impacts data transfer volumes, and consequently, costs.

Calculating Data Transfer Costs

In order to control your data transfer costs, you should first be able to estimate the expected volumes that your environment will generate. With bursting, you need to estimate when spinning additional resources is needed, as well as how much data streams in or out during a peak in demand. Disaster recovery (DR) exemplifies the need to continuously sync your cloud and on-premises environments, costs will be influenced by actions like the frequency of database writes, where data needs to be synced with your public cloud DR environment for an up-to-date backup repository site.

In order to make things a bit more practical, we took a look at two data transfer cost scenarios in AWS.

Database Sync
In principle, data that’s streamed in from the internet or between regions doesn’t accrue costs with AWS. Below you can see how data that is streamed out of your environment can be charged according to different usage tiers. In addition, streaming data between AZs or peered VPCs using public IPs costs $0.01.

Learn about  data transfer costs for cloud deployment


The table below shows a simple case of a master/slave deployments for the production environment part of your database backup, in which case a few GBs are moved out of AWS. As you can see, due to the fact that we are only dealing with a few GBs, the costs are relatively low.

Learn about data transfer costs

Rich Media Streaming

Now, let’s take a look at Netflix. 3Today, Netflix has around 65M subscribers that each consume 45GB on average per month. That’s a total of 2925 Petabytes of data per month. It’s fair to assume that Netflix doesn’t pay list prices for its AWS resources and services. However, even if they pay 1/10 of the cheapest list price ($0.05/10 per GB), the monthly bill would still come out to $14.6M/month. This shows us the significance of data transfer costs.

Final Note
In order to plan, estimate and track your cloud costs, you should first recognize the need to take data transfer costs into consideration when building and operating an application in the cloud. In addition, the high granularity of cloud costs makes continuous control and maintenance quite challenging. In most cases, compute and storage costs are considered first, which is logical seeing as they are definitely thought of as the greatest cost factors. However in some cases, the devil is in the details, leaving you questioning how your massive public cloud bill came to be.

Cloudyn tracks all types of costs in a dynamic cloud environment, including compute, storage and network utilization. And with comprehensive reports and alerts, Cloudyn can help detect and avoid uncontrolled cost spikes that are caused by unexpected data transfers.

Start Optimizing Your Hybrid Cloud Deployment

Start your 20-day free trial today to easily identify greater cloud efficiency opportunities and view your entire AWS deployment in one place. Gain full transparency, accountability, and control.
start free trial


Forgot Password?

No account yet? Register.

0 Flares Twitter 0 Facebook 0 LinkedIn 0 Email -- 0 Flares ×